Scene graph generation for unlabeled data

Abstract

Methods for training and using a scene graph generator for transfer learning are presented. Scene graph generation techniques can decompose domain gaps into individual types of disparities, which can involve appearance, label, and prediction disparities, for example. These disparities can be reduced by aligning corresponding latent and output distributions, at least by using one or more gradient reversal layers (GRLs). Label disparities can be addressed using self-pseudo statistics collected from target data. Self-learning and adversarial techniques based on pseudo statistics can be used to manage these disparities without expensive supervision from real-world datasets.

Description

[0001] Cross Reference to Related Applications

[0002] This PCT application claims priority to U.S. Provisional Patent Application Serial No. 63 / 030,734, filed May 27, 2020, and U.S. Patent Application No. 17 / 226,561, filed April 9, 2021, both of which are incorporated herein in their entirety for all intents and purposes. Background Technology

[0003] A growing number of applications and services rely on techniques such as object recognition and computer vision. To provide accurate results, these techniques often involve the generation of scene graphs and the learning of visual relationships between objects represented in image or video content. Traditional scene graph generation techniques depend on the availability of expensive annotations or labels. Attempting to generate scene graphs without such annotations (e.g., by using unlabeled real data) is a challenging task. Generating synthetic data is a viable alternative and has emerged as a promising direction for extending supervised learning, as the generated labels can be obtained from the data synthesis process. However, reliably performing complex tasks such as scene graph generation for networks or models trained on synthetic data remains challenging when evaluated against real data, due to issues involving performance lag or domain gaps. Attached Figure Description

[0004] Various embodiments according to this disclosure will be described with reference to the accompanying drawings, in which:

[0005] Figure 1A and Figure 1B Images that can be used to generate scene graphs according to at least one embodiment are shown;

[0006] Figure 2 A scene graph generation system according to at least one embodiment is shown;

[0007] Figure 3A and 3B Images and corresponding scene diagrams according to at least one embodiment are shown;

[0008] Figure 4A and 4B Images and corresponding scene diagrams according to at least one embodiment are shown;

[0009] Figure 5 The process for training a scene graph generator according to at least one embodiment is illustrated;

[0010] Figure 6 Components of a system for generating a scene graph according to at least one embodiment are shown;

[0011] Figure 7A Inference and / or training logic is shown;

[0012] Figure 7B Inference and / or training logic is shown;

[0013] Figure 8 An example data center system is shown, according to at least one embodiment;

[0014] Figure 9 A computer system is shown, according to at least one embodiment;

[0015] Figure 10 A computer system is shown, according to at least one embodiment;

[0016] Figure 11 At least portions of a graphics processor are shown, according to one or more embodiments;

[0017] Figure 12 At least portions of a graphics processor are shown, according to one or more embodiments;

[0018] Figure 13 is an example dataflow graph of a high-level compute pipeline, according to at least one embodiment;

[0019] Figure 14 is a system diagram of an example system for training, adapting, instantiating, and deploying machine learning models in a high-level compute pipeline, according to at least one embodiment; and

[0020] Figure 15A and Figure 15B A dataflow graph of a process for training a machine learning model, and an example illustration of a client-server architecture that leverages a pre-trained annotation model to enhance an annotation tool, are shown, according to at least one embodiment. DETAILED DESCRIPTION

[0021] Methods in accordance with various embodiments can provide for the generation and use of scene graphs or other such data representations. In particular, embodiments are directed to scene graph generation for transfer learning. Because conventional scene graph training methods utilize labeled data that is costly or difficult to obtain, it can be advantageous to utilize synthetic data that can be provided labeled with little or no additional cost. Unfortunately, models trained using synthetic data often do not perform well when run on inferences on real data, at least in part due to domain gaps. To overcome this gap problem, scene generation techniques can be utilized that decompose the domain gap into individual types of disparities, such as can involve appearance, label, and prediction disparities. This can include minimizing appearance and prediction disparities by aligning the respective latent and output distributions using a gradient reversal layer. Label disparities can be addressed using self-learning principles, but using pseudo statistics gathered from target data and utilizing a synthetic data generator to produce accurate labels. Pseudo statistics based self-learning and adversarial techniques are used to manage these disparities without the need for expensive supervision from real-world datasets.

[0022] When training a model or network to generate scene graphs, it can be desirable to be able to generate scene graphs that can correspond to inputs of different domains. In this context, a domain can refer to different settings, classes, styles, or spaces, among other such options. For example, a first set of images can include images taken in a first location, such as a metropolitan area in the United States. As shown in example images 100, these images can include particular types of objects of interest to a particular application, such as can include representations of vehicles 102 and pedestrians 104, as well as other such object types that can be present in such locations. These images can also include representations of other types of objects, as well as can include buildings, road signs, billboards, and the like. For labeled data, this image can also include or have associated a set of bounding boxes that illustrate the general locations or regions of the image that correspond to individual instances of these object types of interest. These bounding boxes or other such location identifiers can be used to help train a model or network to recognize these types of objects in similar images. The network can then learn to distinguish these objects with high accuracy when surrounded by buildings and other types of objects that are representative of this general domain or in the metropolitan area of the United States. Figure 1A

[0023] However, it can be possible that it can be desirable to use such a trained model to perform similar classification or inference on data from a different domain. This can include, for example, a domain corresponding to a European alpine region, which can include such as Figure 1B ​The images can include very different types of backgrounds and other objects, as can involve mountains, farm animals, a cabin in the mountains, and other objects that can have very different appearances than those present in the images from the first domain. While there can be similar objects of interest in these images from the second domain, the model or network can not be as accurate in identifying these objects and distinguishing them from surrounding image regions because the model was not trained using images with these representations. To improve the accuracy of such a network (e.g., a scene graph generation network), it can be desirable to train the network using data from the first domain but also using data from the second domain. However, in this example, the data from the second domain is not labeled. As mentioned, traversing and manually annotating these images can be expensive and time consuming, especially for the large number of images that can be needed to accurately train a model.

[0024] In some instances, synthetic images can be generated that include representations of objects of a type of interest. Because the synthetic images can be generated using, for example, scene graphs, these synthetic images can include labels that are known for the type of object included in these synthetic images. It is also possible to synthesize images for a particular domain, such as the European alpine region. Unfortunately, this still results in different domains because there will be images in a first domain (here, a synthetic domain) and images in a second domain (here, a “real” image domain). As such, even training a model on labeled synthetic images of a target domain, the performance can not be accurate enough for real images captured or otherwise obtained for that target domain. As mentioned, this is largely due to domain gaps between domains, which can include both content and appearance gap portions. The appearance gap can represent a difference in visual appearance between real images and synthetic images or images from a different domain. The content gap can represent a difference in placement, number, selection, and type of objects in images and other such options. In at least some embodiments, the appearance gap can be addressed by making the scene photorealistic, such as by using one or more image transformations, using feature alignment, or learning robust representations based on domain randomization. Methods can also be used to address the content gap for image classification.

[0025] Accordingly, methods in accordance with various embodiments can provide scalable methods for training scene graph generation models, networks, or other such methods to address such domain gaps. These methods can further leverage both labeled synthetic data and unlabeled real data. Methods in accordance with various embodiments can provide tractable ways to reduce the gap in appearance and content that can occur between domains, such as between synthetic data and real data. Such methods can also provide for learning of neural networks to generate scene graphs for unlabeled real images. This can include using synthetic data for learning to generate scene graphs (SGs) on real images or images from different domains. Such methods can leverage a generation pipeline that includes pseudo statistics to manage domain gaps. In at least one embodiment, scene graph generation can be learned for various applications or scenarios, such as for computer vision, object recognition, or navigation assistance, among other such options. Applications can be extended to places where synthetic data can be generated with only an original estimate of the target data set distribution.

[0026] As noted above, scene graphs (SGs) can be used in applications such as computer vision and computer graphics to provide an interpretable and structured representation of a scene. Scene graphs can be used to identify or summarize entities in a scene, and reasonable relationships between or among them. Scene graphs can be presented as visual manifestations by inverting graphs. Scene graphs have proven to be beneficial for various applications, including but not limited to image captioning, visual question answering, high-level reasoning tasks, image retrieval, image generation, etc. Much of this existing work on scene graph generation relies on the availability of expensive and limited labeled data sets, such as Visual Genome and Visual Relationship Data Set (VRD). A significant limitation for at least some machine learning applications is the general lack of sufficient labeled data for supervised learning tasks. Synthetic data can provide a viable alternative to this problem, as annotations are essentially free as part of the synthetic process. Synthetic data has been used for various tasks, such as image classification, object detection, semantic segmentation, optical flow modeling, 3D keypoint extraction, object pose estimation, and 3D reconstruction, among others. Synthetic data has also shown to be effective in initializing task networks and for data augmentation. Synthetic data has not been leveraged (at least sufficiently) for scene graph generation and representation of visual relationships.

[0027] In at least one embodiment, it can be assumed that there are gaps in both appearance and content, and these gaps can be extended into individual addressable subcomponents. This approach can help reduce domain gaps for scene graph generation, and these techniques can also be applied to other vision related tasks such as image classification, image segmentation, and object detection, among others. In at least one embodiment, a simulation-to-scene graph (SimS2G) approach can be utilized in which a model learns simulation-to-real scene graph generation using labeled synthetic data and unlabeled real data. The Sim2SG approach can address domain gaps by bounding error of a task (such as scene graph generation) on real data through appearance, prediction, and label (e.g., ground truth) differences between two domains and task error on synthetic data. This approach can minimize appearance and prediction differences by aligning respective latent and output distributions via one or more gradient reversal layers (GRLs). In at least one embodiment, self-learning can be used to address differences on labels. However, self-learning based on pseudo labels can generate inaccurate labels that cause performance issues. Thus, approaches in accordance with various embodiments can collect pseudo statistics from target data and utilize a synthetic data generator to produce accurate labels.

[0028] Figure 2 An example pipeline 200 that can be used for scene graph generation is shown in FIG. 2. As shown, this pipeline can receive input from at least two domains, such as can involve labeled synthetic images 202 and unlabeled “real” or captured images 204. This example pipeline can function in three stages, although other pipelines or implementations can use a single stage or other number of stages. In a first stage, the pipeline takes labeled synthetic images 202 or images from a first domain and processes these images using an encoder 206 to determine relevant features of these images and encode those features into a latent space or representation 208. This pipeline can then process this latent space with a scene graph predictor 210 to attempt to generate an appropriate scene graph for each of those input images. In this example, both the encoder 206 and the scene graph predictor 210 can be neural networks, such as a convolutional neural network (CNN) for the encoder 206 and a generative neural network for the scene graph predictor 210. In this first state, this pipeline can learn to generate scene graphs for labeled synthetic images.

[0029] In the second stage, this pipeline 200 can process unlabelled real images 204. The unlabelled real images 204 are processed using an encoder 206 to determine relevant features and encode those features into a latent space or representation, here a shared representation Z 208 that serves as a single latent space for both labelled synthetic images 202 and unlabelled real images 204. In at least one embodiment, the encoder 206 maps both labelled synthetic images and unlabelled real images to this shared representation Z 208. In at least one embodiment, the encoder can be a ResNet-101, ResNet-50, or VGGNet (e.g., VGG-16) network, among others, and in at least some embodiments, different encoders can be used for different types or categories of input data.

[0030] In the third stage, a scene graph predictor model 210 generates one or more scene graphs from the unlabelled real data in the shared latent space 208. One or more pseudo statistics can then be computed for these scene graphs. In at least one embodiment, these pseudo statistics can represent a high-level statistical description of the scene, which can then be fed back to the generator, here the scene graph predictor 210. These pseudo statistics can then be used to attempt to generate aligned scenes, where the synthetic data more closely mimics the real data, such as where the synthetic images have an appearance that more closely mimics that of real captured images. This aligned data, such as can include generated scene graph data 216, can then be fed to a renderer (for image data) and used as additional data for training the encoder 206 and the generator or scene graph predictor 210 (such as a graph convolutional network). This data can be used to train these networks again for another epoch. After this additional training, the networks can be used to generate another round of aligned data, and this process can continue until an end criterion is met, for example, the networks are converging, a maximum number of training iterations or epochs is reached, or all of the training data for training is used, among other such options.

[0031] In this process, the scene graph prediction network P can be trained on the shared representation Z using synthetic data. Domain gaps can be bridged by aligning features in the representation space Z (for the appearance gap) using a first gradient reversal layer (GRL) 212 and features in the output space (for the content gap) using a second GRL 214, along with a domain discriminator. Using self-learning based on pseudo statistics, the labelled disparity (of the content gap) can be aligned between the two domains. Note that the arrows in the figure correspond to the flow of synthetic data, real data, and / or backpropagation. The GRL can be a feature layer before the final scene graph prediction layer, where the type of discriminator will try to determine the domain of the features, e.g., whether they correspond to the real domain or the synthetic domain.

[0032] Such methods can be beneficial, at least because initial scene graph predictions based on synthetic pre-training will not be accurate, and this inaccuracy is at least partly due to domain disparities. After the initial scene graph predictions, pseudostatistics can be identified, such as the number or layout of objects in the scene, and this can be used to help generate slightly more aligned data in the next iteration or epoch. Better aligned data helps improve the performance of the synthetic prediction network, as well as the performance of predicting or obtaining pseudostatistics from the real data. Over time, this self-learning-based approach can predict data that is increasingly aligned with the real data, causing the network to converge and the synthetic data to be a very accurate representation of the real data.

[0033] As mentioned, in the third state, this synthetic data, which now accurately resembles the real data, can have appearance and predictive features aligned between the two domains. A GRL can be used for both appearance and prediction losses, and the distribution of features is matched at the encoder level and at the output level. Using such a process, appearance alignment will effectively generate features that are either invariant or appearance-invariant in both domains. The prediction loss can also be used to teach the network to be consistent in both domains. Once trained, a set of features classified as "chair" in domain A should also be classified as "chair" in domain B, which helps to align with the prediction loss.

[0034] In at least one embodiment, the SG predictor module can be trained using labels from the source domain. In at least one embodiment, the scene graph can be effectively a set of bounding boxes used as nodes of a graph, where the edges of the graph represent relationships between those objects. The scene graph of a given image I can be labeled as a graph G with nodes o and edges r. Nodes can be bounding boxes b. i ={xmin i ,ymin i ,w i ,h i} and category c i tuple o i = i ,c i >. Relation r is <o i ,p,o j The triple of >, where p is the predicate. In at least one embodiment, scene graph prediction has at least two key components, including an extractor. P-plot predictor and extractor The input space x is mapped to the latent space z, and the predictor P is mapped from the latent space z to the output space y(G). The predicted SG is Networks such as Resnet-101 can be used to implement Furthermore, architectures such as GraphRCNN are used to implement P. These networks can be trained using the following task loss. ​and P: cross-entropy loss for object classification and relation classification, and l1 loss for bounding boxes, as discussed with respect to Figure 2 In at least some embodiments, relationships between objects in a scene are predicted, but not attributes thereof.

[0035] In at least one embodiment, domain gaps between a source domain and a target domain can be analyzed to formulate task error on the real domain (e.g., for SG generation) as a function of task error on the synthetic domain, as well as appearance and content gaps between the domains. Each subcomponent of this domain gap can then be addressed.

[0036] In an example representing a domain gap, two domains will be discussed, including a synthetic domain <x s ,y s > ~ p(x, y) and a real domain <x r ,y r > ~ q(x, y). In this example, y r is unknown. This example also assumes that the synthetic and real domains contain the same classes of objects. The two domains also share similar scenes (e.g., both have driving scenes). However, the joint distribution of scenes (e.g., images) is different in the two domains (i.e., p(x) ≠ q(x)). The label (e.g., ground truth) distribution is also different in the two domains (i.e., p(y) ≠ q(y)). The task (e.g., SG generation) error can then be represented in the synthetic domain as a function of the latent space z:

[0037]

[0038] where e s is the risk defined as e s = |p(y | z) - y s |. Here, p(z) represents the distribution of features, p(y) is the distribution of labels, and p(y | z) is the output distribution. Since the labels y s are accessible, the task error can be minimized. However, the label distribution q(y) of the real domain can not be accessible. Thus, the task error or risk can be tied in the real domain by the following tractable components:

[0039]

[0040] Using the simplification in the above equation, it can be shown that the task error on the real domain is a function of three terms: the task error on the synthetic domain the feature difference between the two domains and the risk difference between the domains For simplicity, the terms and P will be dropped from the continuing discussion. Since features often take appearance into account, the feature difference ∈ a can be called appearance gap, and is shown empirically through ablation. Similarly, analyzing risk difference ∈ c and ablation can be related to content gap. Nonetheless, one approach according to at least one embodiment can attempt to minimize ∈ s , ∈ a , and ∈ c to reduce task error on the real domain. If error ∈ r is reduced to zero on the target domain, then the approach has effectively closed the domain gap. The first term ∈ s can be straightforward to minimize as discussed herein.

[0041] Regarding the second term, appearance gap refers to the difference in appearance of the two domains. This can include, for example, differences in texture, color, light, and reflectance of objects in the scene. To address appearance gap, one approach can attempt to avoid photo-realism in synthetic data, as this can require high-quality assets and significant effort from artists. Thus, one can learn appearance-invariant representations z to avoid appearance bias in the model. However, features z: have both content and appearance components, and aligning these features can be detrimental. To overcome this challenge, one can align the content gap ∈ c of the two domains before aligning the appearance gap ∈ a as discussed herein. This appearance gap ∈ a can be minimized by aligning feature distributions p(z) and q(z) as follows:

[0042] θ * = arg θ min∫(q(z)-p(z))e s dz

[0043] where the fact that p(z)≡q(z) is a sufficient condition to make ∈ a zero is utilized. A gradient reversal layer (GRL) can be used to align distributions p(z) and q(z), as well as classify them with a domain classifier D. The loss on D can be minimized with respect to its own parameters, while simultaneously maximizing with respect to network parameters . This can be performed using a GRL that acts as an identity function during forward propagation and flips the sign of the gradient during backpropagation from D to .

[0044] Regarding the third point, content gap refers to the difference between two domains, including differences in the number of objects and their classes, layouts, poses, and distributions of extensions. This can lead to issues with token offsets or truth value distributions. It is expected that this content gap will be understood in at least some embodiments. Since minimizing ∈ in the current form... c It is unmanageable, therefore the risk difference (e) r -e s It can be reduced as follows:

[0045] arg θ min∈ c =arg θ min∫q(z)(y s -y r )dz+∫q(z)(q(y|z)-p(y|z))dz

[0046] =arg θ min ∈ c，label +∈ c，pred

[0047] We can assume that the risk is positive, and that risk dependence is further divided into labeled differences ∈ c，pred .

[0048] To address the prediction discrepancy ∈ c，pred The output distributions can be aligned for latent codes belonging to the same visual category. This is reasonable because the outputs of scene graph generation models should be identical for the same category in different domains. Terms ∈ can be handled by aligning the output distributions p(y|z) and q(y|z) using the same GRL-based technique discussed earlier. c，pred .

[0049] On the other hand, minimizing ∈ c，label It can be challenging because there may not be a label for the target (e.g., the real) domain (y). r Access to ) . In at least one embodiment, self-learning using pseudostatistics can be used to obtain an estimate of the label distribution for obtaining real data. Pseudo-label-based self-learning may produce inaccurate labels due to prediction inaccuracies when evaluated across different domains. However, the method proposed herein can use a synthetic data generator to produce accurate labels. It can be noted that pseudostatistical methods can be orthogonal to pseudo-label-based self-learning, and the latter can potentially be combined with it.

[0050] In at least one embodiment, the marker alignment ∈ c，label At least in part by targeting the target domain All N scenarios (x r It generates N scene graphs (SGs) to work. For each predicted SG... where k e {1,..., N} and the extractable pseudo statistics ps k . The statistics ps k include a list of objects {o1, o2,..., o j} and relationships {r1, r2,..., r j} between them. Each object o j has a bounding box b j and a class c j associated with it, and each relationship triple r j has two objects and a predicate as described herein. This pseudo statistics can be subsequently used to create a scene using the synthetic data generator. A first object can be selected from ps j according to the class c k of the first object. Then, an asset can be randomly selected from that class and its two-dimensional bounding box b j projected into a three-dimensional scene using camera parameters. If available, camera parameters of the target data can be used, or prudent assumptions can be made that can relate to height or pitch, etc.

[0051] The pose can be chosen intelligently or randomly. For example, for a driving simulation environment, the aspect ratio of the bounding box b j of a car can be used to align it with the lane (0° or 180° rotation) or to rotate it by 90°. The same steps can be repeated for each object in the scene. These relationships can be subsequently used for fine-tuning and correction of object positions and poses.

[0052] For example, in a restaurant simulation environment, there can be different objects visible in the scene, such as tables 302, chairs 304, 306, 308, and laptops 310, 312 visible in the image 300, as shown in Figure 3A At least one embodiment, an “on” relationship can be used to specify that the laptops are placed on the tables. Different assumptions can be made, such as a flat ground plane, all objects reserved on the ground, minimal background (e.g., walls, ground, road, or other objects), and real-world scale. A view of the three-dimensional scene can be subsequently rendered using the synthetic data generator to obtain a synthetic image 300 and ground truth x s , y s . Synthetic data can be generated for all statistics ps k to train the encoding and generation networks at the next epoch. This can be referred to as self-learning, as the prediction of unlabelled targets is used to generate synthetic data similar to the targets, and the model is trained with new data at the next epoch. Figure 3BAn example scene graph 350 that can be used to generate such a scene is shown, indicating here the positional relationships of the chair and laptop with the table 302. Various other scene graphs can likewise be generated for such a scene within the scope of various embodiments.

[0053] Figure 4A and 4B Another example image and scene graph that can be generated according to various embodiments are shown, respectively. In the example image 400, there are three objects represented, including a cylinder 402, a cube 404, and a sphere 406. As shown in the scene graph 450, the positional relationships of each pair of objects in the scene can be included, rather than relative to a single object in the scene. Such a scene graph can contain redundant information, such as that the cylinder is to the right of the cube, and the cube is to the left of the cylinder, but such an approach can provide a number of paths that can be used to generate and check the relationships used to generate the scene from the scene graph. Figure 4A Figure 4B

[0054] In at least one embodiment, the process for training one or more networks, as can be part of a SimS2G pipeline, can assume the availability of an encoder and scene graph predictor, as well as synthetic images, synthetic labels, a data generator, and real images. For a given number of iterations or epochs, a loss function can be attempted to be minimized, including loss terms for appearance gap, content prediction gap, and scene graph generation task. After appropriate network parameters have been adjusted to minimize the loss, appropriate pseudo statistics can be extracted from the predictions, and synthetic data can be generated with labels aligned to real data.

[0055] Figure 5 ​​An example process 500 that can be used to train a scene graph generation model is shown, in accordance with at least one embodiment. This process can utilize a loss function for items having the appearance, content, and SG generation items described above. In this example, a set of labeled synthetic data and a set of unlabeled "real" or non-synthetic data are obtained 502. This data can be input to at least one encoder network, which can extract features of these data sets and encode 504 them into a shared representation, such as a shared latent space. This latent space can be provided 506 as input to a scene graph prediction network, which can predict scene graphs based on features from this latent space. In this example, the features in this latent space can be aligned with the output space of the scene graph prediction network. This can result in an alignment of real and synthetic features, such that synthetic data (e.g., images) generated using the inferred scene graphs can be indistinguishable from real data. As described above, this can also be used to minimize the gap between other domains. The labels of the objects in these scene graphs can also be aligned 510 using self-learning based on pseudo statistics as discussed herein, such that synthetic data can be produced with labels that are aligned with the labels of real data. The scene graph generation model can be trained 512 until an end criterion is met, such as network convergence or a training process for a determined number of epochs (e.g., three epochs). Once trained, the pipeline can be used for different purposes, such as generating synthetic data for simulation or other such purposes. This approach can also be used to generate scene graphs for unlabeled data from different domains. Given data of a real scene without labels for the data, scene graphs can be generated that can be used to synthesize labeled versions of the scene.

[0056] Such an approach can be beneficial for any content creation application, such as for online gaming, where it can be desirable to generate realistic synthetic data. This approach can also be used to automatically build simulation environments, which can be useful for robot or navigation simulation. The approach according to various embodiments is also not limited to scene graphs, but can be used with any related representation of two or more domains, particularly in cases where there is insufficient labeled training data available for one of those domains. Applications that can benefit from such functionality can include area / room scanning, security, space analysis, anomaly detection, and reality simulation, among many others.

[0057] As an example, Figure 6An example network configuration 600 that can be used to provide or generate content is shown. In at least one embodiment, a client device 602 can generate content for a session using components of a content application 604 on the client device 602 and data stored locally on the client device. In at least one embodiment, a content application 624 (e.g., an image generation or editing application) executing on a content server 620 (e.g., a cloud server or edge server) can initiate a session associated with at least the client device 602, as can utilize a session manager and user data stored in a user database 634, and can cause content 632 to be determined by a content manager 626, and rendered using a rendering engine if this type of content or platform requires it, and sent to the client device 602 using an appropriate transport manager 622 for transmission by download, streaming, or other such transmission channel. In at least one embodiment, this content 632 can include assets that a rendering engine can use to render a scene based on a determined scene graph. The content server 620 can also include one or more training modules 630 for training a scene graph generation component, network, or pipeline 628. In at least one embodiment, a client device 602 receiving this content can provide the content to a corresponding content application 604, which can also or instead include a rendering engine if needed, for rendering at least some of the content for presentation via the client device 602, such as image or video content through a display 606 and audio, such as sound and music, through at least one audio playback device 608, such as speakers or headphones. For example, for live video content captured by one or more cameras, such a rendering engine can not be needed unless to enhance the video content in some way.

[0058] In at least one embodiment, at least some of this content can already be stored on the client device 602, rendered on the client device 602, or accessible to the client device 602, such that no transmission over the network 640 is needed for at least this portion of the content, such as where the content can have been previously downloaded or stored locally on a hard drive or optical disc. In at least one embodiment, a transmission mechanism such as a data stream can be used to transmit this content from the content server 620 or content database 634 to the client device 602. In at least one embodiment, at least a portion of this content can be obtained or streamed from another source, such as a third party content service 660, which can also include a content application 662 for generating or providing content. In at least one embodiment, portions of this functionality can be performed using multiple computing devices or multiple processors within one or more computing devices, such as a combination that can include a CPU and a GPU.

[0059] In at least one embodiment, content application 624 includes a content manager 626 that can determine or analyze content prior to transmission of that content to client device 602. In at least one embodiment, content manager 626 can also include or work in conjunction with other components that can generate, modify, or enhance content to be provided. In at least one embodiment, this can include a rendering engine for rendering image or video content. In at least one embodiment, an image, video, or scene graph generation component 628 can be used to generate a scene graph that can be used by content application 624 on a server or content application 604 on a client device to generate image, video, or other media content. In at least one embodiment, an enhancement component 630, which can also include a neural network, can perform one or more enhancements to that content as discussed and suggested herein. In at least one embodiment, content manager 626 can cause that content to be transmitted to client device 602. In at least one embodiment, content application 604 on client device 602 can also include components such as a rendering engine, a scene graph generator 612, and a content generation module, such that any or all of that functionality can additionally or alternatively be performed on client device 602. In at least one embodiment, content application 662 on third party content service system 660 can also include such functionality. In at least one embodiment, a location where at least some of that functionality is performed can be configurable, or can depend on factors such as a type of client device 602 or availability of a network connection with sufficient bandwidth. In at least one embodiment, a system for content generation can include any suitable combination of hardware and software in one or more locations. In at least one embodiment, generated image or video content at one or more resolutions can also be provided to or made available to other client devices 650, such as for download or streaming from a media source that stores a copy of that image or video content. In at least one embodiment, this can include transmission of game content for a multiplayer game, where different client devices can display that content at different resolutions, including one or more super- resolutions.

[0060] In this example, the client devices can include any appropriate computing devices, such as can include desktop computers, notebook computers, edge devices, set-top boxes, streaming devices, gaming consoles, smartphones, tablet computers, VR headsets, AR eyewear, wearable computers, or smart televisions. Each client device can submit requests across at least one wired or wireless network, such as can include the Internet, an Ethernet network, a local area network (LAN), or a cellular network, among other such options. In this example, the requests can be submitted to an address associated with a cloud provider, which can operate or control one or more electronic resources in a cloud provider environment, such as can include a data center or a server farm. In at least one embodiment, the requests can be received or processed by at least one edge server located at the edge of the network and outside of at least one security layer associated with the cloud provider environment. In this way, latency can be reduced by causing the client devices to interact with servers that are closer in distance, while also improving security of the resources in the cloud provider environment.

[0061] In at least one embodiment, such a system can be used to perform graphics rendering operations. In other embodiments, such a system can be used for other purposes, such as to provide image or video content for testing or validating autonomous machine applications, or to perform deep learning operations. In at least one embodiment, such a system can be implemented using edge devices, or can incorporate one or more virtual machines (VMs). In at least one embodiment, such a system can be implemented at least partially in a data center or at least partially using cloud computing resources.

[0062] Inference and training logic

[0063] Figure 7A Inference and / or training logic 715 are used to perform inferencing and / or training operations associated with one or more embodiments. In at least one embodiment, as shown in FIG. 7, inference and / or training logic 715 can be used in place of or in conjunction with inference and / or training logic 715 described above. Figure 7A and / or Figure 7B Details regarding inference and / or training logic 715 are provided below in conjunction with FIG. 7.

[0064] In at least one embodiment, inference and / or training logic 715 can include, without limitation, code and / or data storage 701 for storing forward and / or output weights and / or input / output data, and / or other parameters of neurons or layers of a neural network configured in aspects of one or more embodiments that are trained and / or used for inferencing. In at least one embodiment, training logic 715 can include or be coupled to code and / or data storage 701 for storing graph code or other software to control timing and / or order, in which weight and / or other parameter information is loaded to configure logic, including integer and / or floating point units (collectively, arithmetic logic units (ALUs)). In at least one embodiment, code, such as graph code, loads weight or other parameter information into processor ALUs based on an architecture of a neural network to which the code corresponds. In at least one embodiment, code and / or data storage 701 stores weight parameters and / or input / output data of each layer of a neural network trained or used in conjunction with one or more embodiments during forward propagation of input / output data and / or weight parameters during training and / or inferencing using aspects of one or more embodiments. In at least one embodiment, any portion of code and / or data storage 701 can be included with other on-chip or off-chip data storage, including a processor’s Ll, L2, or L3 cache or system memory.

[0065] In at least one embodiment, any portions of code and / or data storage 701 can be internal or external to one or more processors or other hardware logic devices or circuits. In at least one embodiment, code and / or data storage 701 can be cache memory, dynamic random addressable memory (“DRAM”), static random addressable memory (“SRAM”), non-transitory memory such as flash memory, or other storage. In at least one embodiment, a choice of whether code and / or data storage 701 is internal or external to a processor, e.g., or comprised of DRAM, SRAM, flash, or some other storage type, can depend on available storage space on-chip or off-chip, latency requirements of training and / or inferencing functions being performed, batch size of data used in inferencing and / or training of a neural network, or some combination of these factors. In at least one embodiment, inferencing and / or training logic 715 can include, without limitation, code and / or data storage 705 to store backward and / or output weight and / or input / output data corresponding to neurons or layers of a neural network being trained and / or used for inferencing in aspects of one or more embodiments. In at least one embodiment, code and / or data storage 705 stores weight parameters and / or input / output data of each layer of a neural network trained or used in conjunction with one or more embodiments during backpropagation of input / output data and / or weight parameters during training and / or inferencing using aspects of one or more embodiments. In at least one embodiment, training logic 715 can include or be coupled to code and / or data storage 705 to store graph code or other software to control timing and / or order, wherein weight and / or other parameter information is loaded to configure logic, including integer and / or floating point units (collectively, arithmetic logic units (ALUs)).

[0066] In at least one embodiment, code such as graph code loads weight or other parameter information into processor ALUs based on an architecture of a neural network that code corresponds to. In at least one embodiment, any portion of code and / or data storage 705 can be included with other on-chip or off-chip data storage, including a processor’s L1, L2, or L3 cache or system memory. In at least one embodiment, any portion of code and / or data storage 705 can be internal or external to one or more processors or other hardware logic devices or circuits. In at least one embodiment, code and / or data storage 705 can be cache memory, DRAM, SRAM, non-transitory memory (e.g., flash memory), or other storage. In at least one embodiment, a choice of whether code and / or data storage 705 is internal or external to a processor, e.g., is made up of DRAM, SRAM, flash memory, or some other storage type, depends on whether available storage is on-chip or off-chip, latency requirements of training and / or inferencing functions being performed, batch size of data being used in inferencing and / or training of a neural network, or some combination of these factors.

[0067] In at least one embodiment, code and / or data storage 701 and code and / or data storage 705 can be separate storage structures. In at least one embodiment, code and / or data storage 701 and code and / or data storage 705 can be the same storage structure. In at least one embodiment, code and / or data storage 701 and code and / or data storage 705 can be partially the same storage structure and partially separate storage structures. In at least one embodiment, any portion of code and / or data storage 701 and code and / or data storage 705 can be included with other on-chip or off-chip data storage, including a processor’s L1, L2, or L3 cache or system memory.

[0068] In at least one embodiment, inference and / or training logic 715 can include, without limitation, one or more arithmetic logic unit(s) (“ALUs”) 710 (including integer and / or floating point units) for performing logical and / or mathematical operations, based at least in part on training and / or inference code (e.g., graph code) or instructions by which they are instructed to operate, whose results can be activations (e.g., output values from layers or neurons within a neural network) stored in an activation storage 720, which are functions of input / output and / or weight parameter data stored in code and / or data storage 701 and / or code and / or data storage 705. In at least one embodiment, activations stored in activation storage 720 are generated by ALU 710 executing linear algebraic and / or matrix-based mathematics in response to executing instructions or other code, where weight values stored in code and / or data storage 705 and / or code and / or data storage 701 are used as operands along with other values, such as bias values, gradient information, momentum values, or other parameters or hyperparameters, any or all of which can be stored in code and / or data storage 705 or code and / or data storage 701 or other on-chip or off-chip storage.

[0069] In at least one embodiment, one or more ALUs 710 are included in one or more processors or other hardware logic devices or circuits, while in another embodiment one or more ALUs 710 can be external to a processor or other hardware logic device or circuit using them (e.g., a co-processor). In at least one embodiment, one or more ALUs 710 can be included within execution units of a processor, or otherwise included in a group of ALUs accessible by execution units of a processor, which can be within a same processor or distributed across different types of processors (e.g., central processing units, graphics processing units, fixed function units, etc.). In at least one embodiment, code and / or data storage 701, code and / or data storage 705, and activation storage 720 can be on a same processor or other hardware logic device or circuit, while in another embodiment they can be on different processors or other hardware logic devices or circuits, or some combination of same and different processors or other hardware logic devices or circuits. In at least one embodiment, any portion of activation storage 720 can be included with other on-chip or off-chip data storage, including L1, L2, or L3 cache of a processor or system memory. Moreover, inference and / or training code can be stored with other code accessible to a processor or other hardware logic or circuit, and can be fetched and / or processed using fetch, decode, schedule, execution, exit, and / or other logic circuits of a processor. In at least one embodiment, activation storage 720 can be cache memory, DRAM, SRAM, non-transitory memory (e.g., flash), or other storage. In at least one embodiment, activation storage 720 can be entirely or partially internal or external to one or more processors or other logic circuits. In at least one embodiment, whether activation storage 720 is internal or external to a processor, or contains DRAM, SRAM, flash, or other storage types, can be chosen depending on storage available on-chip or off-chip, latency requirements of training and / or inference functions, batch size of data used in inference and / or training neural networks, or some combination of these factors.

[0070] In at least one embodiment, inference and / or training logic 715 shown in FIG. 7A can be used with Figure 7A application specific integrated circuits (“ASICs”), such as Tensorflow® Processing Units (TPUs) from Google, inference processing units (IPUs) from Graphcore® AI processing units (APUs) from Advanced Micro Devices, Inc. (AMD), inference processing units (IPUs) from Graphcore® AI TM Systems, or Intel® Nervana® processors (e.g., “Lake Crest”) from Intel Corp. In at least one embodiment, inference and / or training logic 715 can be used with other hardware-accelerated machine learning processors or accelerators. Figure 7A ​The illustrated inference and / or training logic 715 can be used in combination with central processing unit (“CPU”) hardware, graphics processing unit (“GPU”) hardware, or other hardware (e.g., field programmable gate arrays (“FPGAs”)).

[0071] Figure 7B Inference and / or training logic 715 is shown in FIG. 7, according to at least one or more embodiments. In at least one embodiment, inference and / or training logic 715 can include, without limitation, hardware logic, in which computing resources are dedicated or otherwise uniquely used in conjunction with weight values or other information corresponding to one or more layers of neurons within a neural network. In at least one embodiment, inference and / or training logic 715 includes, without limitation, hardware logic, in which computing resources are dedicated or otherwise uniquely used in conjunction with weight values or other information corresponding to one or more layers of neurons within a neural network. Figure 7B Inference and / or training logic 715 shown in FIG. 7 can be used in combination with application-specific integrated circuit (ASIC) such as Tensor Processing Units from Google, Tensor Processing Units (TPUs) from Google, TM Inference Processing Units (IPUs) from Graphcore, or “Lake Crest”) processors from Intel Corp. In at least one embodiment, inference and / or training logic 715 includes, without limitation, hardware logic, in which computing resources are dedicated or otherwise uniquely used in conjunction with weight values or other information corresponding to one or more layers of neurons within a neural network. Figure 7B Inference and / or training logic 715 shown in FIG. 7 can be used in combination with central processing unit (CPU) hardware, graphics processing unit (GPU) hardware, or other hardware (e.g., field programmable gate arrays (FPGAs)). In at least one embodiment, inference and / or training logic 715 includes, without limitation, code and / or data storage 701 and code and / or data storage 705, which can be used to store code (e.g., graph code), weight values, and / or other information, including bias values, gradient information, momentum values, and / or other parameter or hyperparameter information. In Figure 7B In at least one embodiment illustrated in FIG. 7, each of code and / or data storage 701 and code and / or data storage 705 is associated with a dedicated computing resource (e.g., computing hardware 702 and computing hardware 706, respectively). In at least one embodiment, each of computing hardware 702 and computing hardware 706 includes one or more ALUs that perform mathematical functions (e.g., linear algebraic functions) only on information stored in code and / or data storage 701 and code and / or data storage 705, respectively, the results of which are stored in activation storage 720.

[0072] In at least one embodiment, each of code and / or data stores 701 and 705 and corresponding compute hardware 702 and 706, respectively, correspond to different layers of a neural network, such that activations resulting from one “storage / compute pair 701 / 702” of code and / or data store 701 and compute hardware 702 provide input to the next “storage / compute pair 705 / 706” of code and / or data store 705 and compute hardware 706 in order to reflect the conceptual organization of a neural network. In at least one embodiment, each storage / compute pair 701 / 702 and 705 / 706 can correspond to more than one neural network layer. In at least one embodiment, additional storage / compute pairs (not shown) can be included in inference and / or training logic 715 after or in parallel with storage compute pairs 701 / 702 and 705 / 706.

[0073] Data center

[0074] Figure 8 An example data center 800 that can use at least one embodiment is shown. In at least one embodiment, data center 800 includes a data center infrastructure layer 810, a framework layer 820, a software layer 830 and an application layer 840.

[0075] In at least one embodiment, as shown Figure 8 Data center infrastructure layer 810 can include resource orchestrators 812, grouped computing resources 814 and node computing resources (“node C.R.s”) 816(1)-816(N), where “N” represents any positive integer. In at least one embodiment, node C.R.s 816(1)-816(N) can include, but are not limited to, any number of central processing units (“CPUs” or “processors”) including accelerators, field programmable gate arrays (FPGAs), graphics processors, etc., memory devices (e.g., dynamic random access memory), storage devices (e.g., solid state or disk drives), network input / output (“NWI / O”) devices, network switches, virtual machines (“VMs”), power modules and cooling modules, etc. In at least one embodiment, one or more of node C.R.s 816(1)-816(N) can be a server having one or more of above-described computing resources.

[0076] In at least one embodiment, the grouped computing resource 814 may include individual groups (not shown) of node CRs housed within one or more racks, or a plurality of racks (also not shown) housed within data centers in various geographic locations. The individual groups of node CRs within the grouped computing resource 814 may include computing, networking, memory, or storage resources that can be configured or allocated to support groups of one or more workloads. In at least one embodiment, several node CRs, including CPUs or processors, may be grouped within one or more racks to provide computing resources to support one or more workloads. In at least one embodiment, the one or more racks may also include any number of power modules, cooling modules, and network switches, in any combination.

[0077] In at least one embodiment, resource coordinator 812 may configure or otherwise control one or more nodes CR816(1)-816(N) and / or grouped computing resources 814. In at least one embodiment, resource coordinator 812 may include a Software Design Infrastructure (“SDI”) management entity for data center 800. In at least one embodiment, resource coordinator 108 may include hardware, software, or some combination thereof.

[0078] In at least one embodiment, such as Figure 8 As shown, framework layer 820 includes a job scheduler 822, a configuration manager 824, a resource manager 826, and a distributed file system 828. In at least one embodiment, framework layer 820 may include a framework of software 832 supporting software layer 830 and / or one or more applications 842 supporting application layer 840. In at least one embodiment, software 832 or application 842 may respectively include web-based service software or applications, such as services or applications provided by Amazon Web Services, Google Cloud, and Microsoft Azure. In at least one embodiment, framework layer 820 may be, but is not limited to, a free and open-source software web application framework, such as Apache Spark, which can utilize distributed file system 828 for large-scale data processing (e.g., "big data"). TM(hereinafter “Spark”). In at least one embodiment, job scheduler 832 can include a Spark driver to facilitate scheduling of workloads supported by various layers of data center 800. In at least one embodiment, configuration manager 824 can be capable of configuring different layers, such as software layer 830 and framework layer 820 including Spark and a distributed file system 828 for supporting large-scale data processing. In at least one embodiment, resource manager 826 can be capable of managing clustered or grouped computing resources mapped to or allocated for supporting distributed file system 828 and job scheduler 822. In at least one embodiment, clustered or grouped computing resources can include grouped computing resources 814 on data center infrastructure layer 810. In at least one embodiment, resource manager 826 can coordinate with resource orchestrator 812 to manage these mapped or allocated computing resources.

[0079] In at least one embodiment, software 832 included in software layer 830 can include software used by at least portions of node C.R.s 816(1)-816(N), grouped computing resources 814, and / or distributed file system 828 of framework layer 820. One or more types of software can include, but are not limited to, Internet web page search software, email virus scanning software, database software, and streaming video content software.

[0080] In at least one embodiment, one or more applications 842 included in application layer 840 can include one or more types of applications used by at least portions of node C.R.s 816(1)-816(N), grouped computing resources 814, and / or distributed file system 828 of framework layer 820. One or more types of applications can include, but are not limited to, any number of genomics applications, cognitive computing and machine learning applications, including training or inferencing software, machine learning framework software (e.g., PyTorch, TensorFlow, Caffe, etc.), or other machine learning applications used in conjunction with one or more embodiments.

[0081] In at least one embodiment, any of configuration manager 824, resource manager 826, and resource orchestrator 812 can implement any number and type of self-modifying actions based on any number and type of data acquired in any technically feasible fashion. In at least one embodiment, self-modifying actions can relieve data center operators of data center 800 from making possibly poor configuration decisions and can avoid underutilized and / or poorly performing portions of a data center.

[0082] In at least one embodiment, data center 800 can include tools, services, software or other resources to train one or more machine learning models or use one or more machine learning models to predict or infer information according to one or more embodiments described herein. For example, in at least one embodiment, a machine learning model can be trained according to a neural network architecture by computing weight parameters using software and computing resources described above with respect to data center 800. In at least one embodiment, using resources described above with respect to data center 800, a trained machine learning model corresponding to one or more neural networks can be used to infer or predict information using weight parameters computed by one or more training techniques described herein.

[0083] In at least one embodiment, data center can use CPUs, application specific integrated circuits (ASICs), GPUs, FPGAs, or other hardware to perform training and / or inference using resources described above. Moreover, one or more software and / or hardware resources described above can be configured as a service to allow users to train or perform information inference such as image recognition, speech recognition, or other artificial intelligence services.

[0084] Inference and / or training logic 715 are used to perform inferencing and / or training operations associated with one or more embodiments. Details regarding inference and / or training logic 715 are provided below in conjunction with FIGS. 7A and / or 7B. In at least one embodiment, inference and / or training logic 715 can be used in system 700 for inferencing or predicting operations based, at least in part, on weight parameters calculated using neural network training operations, neural network functions and / or architectures, or neural network use cases described herein. Figure 7A And / or Figure 7B Details regarding inference and / or training logic 715 are provided below in conjunction with FIGS. 7A and / or 7B. In at least one embodiment, inference and / or training logic 715 can be used in system 700 for inferencing or predicting operations based, at least in part, on weight parameters calculated using neural network training operations, neural network functions and / or architectures, or neural network use cases described herein. Figure 8

[0085] Such components can be used to train one or more generative models to be able to generate scene graphs for unlabeled data, as well as generate scenes based on those scene graphs that are indistinguishable from real scenes.

[0086] Computer system

[0087] Figure 9 ​This is a block diagram illustrating an exemplary computer system according to at least one embodiment. The exemplary computer system may be a system of interconnected devices and components, a system-on-a-chip (SOC), or some combination thereof formed with a processor, which may include an execution unit to execute instructions. In at least one embodiment, according to this disclosure, such as the embodiments described herein, computer system 900 may include, but is not limited to, components such as processor 902, whose execution unit includes logic to execute algorithms for process data. In at least one embodiment, computer system 900 may include a processor, such as one available from Intel Corporation of Santa Clara, California. Processor family, Xeon TM , XScale TM and / or StrongARM TM , Core TM or Nervana TM A microprocessor may be used, although other systems (including PCs, engineering workstations, set-top boxes, etc.) with other microprocessors may also be used. In at least one embodiment, computer system 900 may execute a version of the Windows operating system available from Microsoft Corporation of Redmond, Washington, although other operating systems (such as UNIX and Linux), embedded software, and / or graphical user interfaces may also be used.

[0088] The embodiments can be used in other devices, such as handheld devices and embedded applications. Some examples of handheld devices include cellular phones, Internet Protocol (IP) devices, digital cameras, personal digital assistants (“PDAs”), and handheld PCs. In at least one embodiment, the embedded application may include a microcontroller, a digital signal processor (“DSP”), a system-on-a-chip (SoC), a network computer (“NetPC”), a set-top box, a network hub, a wide area network (“WAN”) switch, or any other system that can execute one or more instructions according to at least one embodiment.

[0089] In at least one embodiment, computer system 900 can include, without limitation, a processor 902 that can include, without limitation, one or more execution units 908 to perform, e.g., machine learning model training and / or inference, in accordance with techniques described herein. In at least one embodiment, computer system 900 is a single processor desktop or server system, but in another embodiment, computer system 900 can be a multiprocessor system. In at least one embodiment, processor 902 can include, without limitation, a complex instruction set computer (“CISC”) microprocessor, a reduced instruction set computing (“RISC”) microprocessor, a very long instruction word (“VLIW”) microprocessor, a processor implementing a combo of instruction sets, or any other processor device, such as a digital signal processor. In at least one embodiment, processor 902 can be coupled to a processor bus 910 that can transmit data signals between processor 902 and other components in computer system 900.

[0090] In at least one embodiment, processor 902 can include, without limitation, a level 1 (“L1”) internal cache memory (“cache”) 904. In at least one embodiment, processor 902 can have a single internal cache or multiple levels of internal caches. In at least one embodiment, cache memory can reside in the processor 902’s external. Other embodiments can include a combination of internal and external caches depending on the specific implementation and requirements. In at least one embodiment, register file 906 can store different types of data within various registers including, without limitation, integer registers, floating point registers, status registers, and instruction pointer registers.

[0091] In at least one embodiment, execution unit(s) 908, including, without limitation, logic to perform integer and floating-point operations, also reside in processor 902. In at least one embodiment, processor 902 can also include a microcode (“ucode”) read-only memory (“ROM”) that stores microcode for certain macro instructions. In at least one embodiment, execution unit(s) 908 can include logic to handle a packed instruction set 909. In at least one embodiment, by including the packed instruction set 909 in a general-purpose processor’s instruction set, along with associated circuitry to execute the instructions, the processor 902 can be used in instances in which data objects, including, without limitation, 128 bytes, are useful. In at least one embodiment, by using the processor’s data

[0092] In at least one embodiment, execution unit 908 can also be used in microcontrollers, embedded processors, graphics devices, DSPs, and other types of logic circuits. In at least one embodiment, computer system 900 can include, but not limited to, memory 920. In at least one embodiment, memory 920 can be implemented as a Dynamic Random Access Memory (“DRAM”) device, a Static Random Access Memory (“SRAM”) device, a flash memory device, or other memory device. In at least one embodiment, memory 920 can store data and / or instructions that can be executed by processor 902 as represented by data signals.

[0093] In at least one embodiment, a system logic chip can be coupled to processor bus 910 and memory 920. In at least one embodiment, system logic chip can include, without limitation, a memory controller hub (“MCH”) 916, and processor 902 can communicate with MCH 916 via processor bus 910. In at least one embodiment, MCH 916 can provide a high bandwidth memory path 918 to memory 920 for instruction and data storage and for storage of graphics commands, data, and textures. In at least one embodiment, MCH 916 can direct data signals between processor 902, memory 920, and other components in computer system 900, and can translate virtual addresses into physical addresses, and vice versa. In at least one embodiment, MCH 916 can also include a display controller to communicate with a display device 922 via an Accelerated Graphics Port (“AGP”) interconnect 914. In at least one embodiment, AGP interconnect 914 can provide a high bandwidth, low latency

[0094] In at least one embodiment, computer system 900 can use system I / O 922, which is a proprietary hub interface bus to couple MCH 916 to I / O controller hub (“ICH”) 930. In at least one embodiment, ICH 930 can provide direct connection to some I / O devices and can be connected to other devices via a local I / O bus. In at least one embodiment, local I / O bus can include, without limitation, a high-speed I / O bus for connecting peripheral devices to memory 920, chipset, and processor 902. Examples can include, without limitation, audio controller 929, firmware hub (“Flash BIOS”) 928, wireless transceiver 926, data storage 924, legacy I / O controller 923 containing user input and keyboard interfaces, serial expansion port 927 (e.g., Universal Serial Bus (USB) port), and network controller 934. Data storage 924 can include a hard disk drive, a floppy disk drive, a CD-ROM device, a flash memory device, or other mass storage device.

[0095] In at least one embodiment, Figure 9 Systems including interconnected hardware devices or “chips” are shown, while in other embodiments, Figure 9 An exemplary system on a chip (SoC) can be shown. In at least one embodiment, devices can be interconnected with proprietary interconnects, standardized interconnects (e.g., PCIe), or some combination thereof. In at least one embodiment, one or more components of computer system 900 are interconnected using compute express link (CXL) interconnects.

[0096] Inference and / or training logic 715 are used to perform inferencing and / or training operations associated with one or more embodiments. Details regarding inference and / or training logic 715 are provided below in conjunction with FIGS. 7 A and / or 7B. In at least one embodiment, inference and / or training logic 715 can be used in Figure 7A and / or Figure 7B Details regarding inference and / or training logic 715 are provided below in conjunction with FIGS. 7 A and / or 7B. In at least one embodiment, inference and / or training logic 715 can be used in Figure 9 neural network training operation, neural network function, and / or architecture or neural network use case described herein.

[0097] Such components can be used to train one or more generative models to be able to generate scene graphs for unlabeled data, as well as generate scenes based on those scene graphs that are indistinguishable from real scenes.

[0098] Figure 10is a block diagram illustrating an electronic device 1000 for utilizing a processor 1010, in accordance with at least one embodiment. In at least one embodiment, electronic device 1000 can be, for example and without limitation, a notebook, a tower server, a rack server, a blade server, a laptop, a desktop, a tablet, a mobile device, a phone, an embedded computer, or any other suitable electronic device.

[0099] In at least one embodiment, system 1000 can include, without limitation, a processor 1010 communicatively coupled to any suitable number or kind of components, peripherals, modules, or devices. In at least one embodiment, processor 1010 is coupled using a bus or interface such as an IMB, a System Management Bus (“SMBus”), a Low Pin Count (LPC) bus, a Serial Peripheral Interface (“SPI”), a High Definition Audio (“HDA”) bus, a Serial Advanced Technology Attachment (“SATA”) bus, a Universal Serial Bus (“USB”) (versions 1, 2, 3), or a Universal Asynchronous Receiver / Transmitter (“UART”) bus. In at least one embodiment, processor 1010 can include, without limitation, a single-core processor, multicore processor, or any other suitable processor. Figure 10 In at least one embodiment, system 1000 can include, without limitation, a processor 1010 communicatively coupled to any suitable number or kind of components, peripherals, modules, or devices. In at least one embodiment, processor 1010 is coupled using a bus or interface such as an IMB, a System Management Bus (“SMBus”), a Low Pin Count (LPC) bus, a Serial Peripheral Interface (“SPI”), a High Definition Audio (“HDA”) bus, a Serial Advanced Technology Attachment (“SATA”) bus, a Universal Serial Bus (“USB”) (versions 1, 2, 3), or a Universal Asynchronous Receiver / Transmitter (“UART”) bus. In at least one embodiment, processor 1010 can include, without limitation, a single-core processor, multicore processor, or any other suitable processor. Figure 10 In at least one embodiment, system 1000 can include, without limitation, a processor 1010 communicatively coupled to any suitable number or kind of components, peripherals, modules, or devices. In at least one embodiment, processor 1010 is coupled using a bus or interface such as an IMB, a System Management Bus (“SMBus”), a Low Pin Count (LPC) bus, a Serial Peripheral Interface (“SPI”), a High Definition Audio (“HDA”) bus, a Serial Advanced Technology Attachment (“SATA”) bus, a Universal Serial Bus (“USB”) (versions 1, 2, 3), or a Universal Asynchronous Receiver / Transmitter (“UART”) bus. In at least one embodiment, processor 1010 can include, without limitation, a single-core processor, multicore processor, or any other suitable processor. Figure 10 In at least one embodiment, system 1000 can include, without limitation, a processor 1010 communicatively coupled to any suitable number or kind of components, peripherals, modules, or devices. In at least one embodiment, processor 1010 is coupled using a bus or interface such as an IMB, a System Management Bus (“SMBus”), a Low Pin Count (LPC) bus, a Serial Peripheral Interface (“SPI”), a High Definition Audio (“HDA”) bus, a Serial Advanced Technology Attachment (“SATA”) bus, a Universal Serial Bus (“USB”) (versions 1, 2, 3), or a Universal Asynchronous Receiver / Transmitter (“UART”) bus. In at least one embodiment, processor 1010 can include, without limitation, a single-core processor, multicore processor, or any other suitable processor. Figure 10 In at least one embodiment, system 1000 can include, without limitation, a processor 1010 communicatively coupled to any suitable number or kind of components, peripherals, modules, or devices. In at least one embodiment, processor 1010 is coupled using a bus or interface such as an IMB, a System Management Bus (“SMBus”), a Low Pin Count (LPC) bus, a Serial Peripheral Interface (“SPI”), a High Definition Audio (“HDA”) bus, a Serial Advanced Technology Attachment (“SATA”) bus, a Universal Serial Bus (“USB”) (versions 1, 2, 3), or a Universal Asynchronous Receiver / Transmitter (“UART”) bus. In at least one embodiment, processor 1010 can include, without limitation, a single-core processor, multicore processor, or any other suitable processor.

[0100] In at least one embodiment, system 1000 can include, without limitation, a processor 1010 communicatively coupled to any suitable number or kind of components, peripherals, modules, or devices. In at least one embodiment, processor 1010 is coupled using a bus or interface such as an IMB, a System Management Bus (“SMBus”), a Low Pin Count (LPC) bus, a Serial Peripheral Interface (“SPI”), a High Definition Audio (“HDA”) bus, a Serial Advanced Technology Attachment (“SATA”) bus, a Universal Serial Bus (“USB”) (versions 1, 2, 3), or a Universal Asynchronous Receiver / Transmitter (“UART”) bus. In at least one embodiment, processor 1010 can include, without limitation, a single-core processor, multicore processor, or any other suitable processor. Figure 10 may include a display 1024, a touchscreen 1025, a touchpad 1030, a near field communication unit (“NFC”) 1045, a sensor hub 1040, a thermal sensor 1046, an express chipset (“EC”) 1035, a trusted platform module (“TPM”) 1038, a BIOS / firmware / flash memory (“BIOS, FW Flash”) 1022, a DSP 1060, a drive 1020 (such as a solid state disk (“SSD”) or a hard disk drive (“HDD”)), a wireless local area network unit (“WLAN”) 1050, a Bluetooth unit 1052, a wireless wide area network unit (“WWAN”) 1056, a global positioning system (GPS) 1055, a camera (“USB3.0 camera”) 1054 (such as a USB3.0 camera), and / or a low power double data rate (“LPDDR”) memory unit (“LPDDR3”) 1015 implemented in, for example, an LPDDR3 standard. These components can each be implemented in any suitable manner.

[0101] In at least one embodiment, other components can be communicatively coupled to processor 1010 by components described above. In at least one embodiment, accelerometer 1041, ambient light sensor (“ALS”) 1042, compass 1043, and gyroscope 1044 can be communicatively coupled to sensor hub 1040. In at least one embodiment, thermal sensor 1039, fan 1037, keyboard 1036, and touchpad 1030 can be communicatively coupled to EC 1035. In at least one embodiment, speaker 1063, headphones 1064, and microphone (“mic”) 1065 can be communicatively coupled to audio unit (“audio codec and class D amplifier”) 1062, which in turn can be communicatively coupled to DSP 1060. In at least one embodiment, audio unit 1062 can include, for example and without limitation, an audio coder / decoder (“codec”) and a class D amplifier. In at least one embodiment, SIM card (“SIM”) 1057 can be communicatively coupled to WWAN unit 1056. In at least one embodiment, components such as WLAN unit 1050 and Bluetooth unit 1052, as well as WWAN unit 1056, can be implemented in a next generation form factor (NGFF).

[0102] Inference and / or training logic 715 are used to perform inferencing and / or training operations associated with one or more embodiments. Details regarding inference and / or training logic 715 are provided below in conjunction with FIGS. 7 and 8. Figure 7A and / or Figure 7B Details regarding inference and / or training logic 715 are provided below in conjunction with FIGS. 7 and 8. Figure 10 Inference operations and / or training operations performed by inference and / or training logic 715, in conjunction with one or more other components of system 700 described herein, can be used to perform inferencing and / or training operations associated with one or more embodiments. For example, in at least one embodiment, inference and / or training logic 715 can be used to perform inferencing and / or training operations associated with a neural network.

[0103] Such components can be used to train one or more generative models to be able to generate scene graphs for unlabeled data, as well as generate scenes based on those scene graphs that are indistinguishable from real scenes.

[0104] Figure 11 FIG. 11 is a block diagram of a processing system, in accordance with at least one embodiment. In at least one embodiment, system 1100 includes one or more processors 1102 and one or more graphics processors 1108, and can be a single processor desktop system, a multiprocessor workstation system, or a server system having many processors 1102 or processor cores 1107. In at least one embodiment, system 1100 is a processing platform incorporated within a system on a chip (SoC) integrated circuit for use in mobile, handheld, or embedded devices.

[0105] In at least one embodiment, system 1100 can include or be incorporated into a game console, including a game and media console, a mobile game console, a handheld game console, or an online game console. In at least one embodiment, system 1100 is a mobile phone, a smart phone, a tablet device, or a mobile internet device. In at least one embodiment, system 1100 can also include or be incorporated into a television or set-top box device having one or more processors 1102 and a graphical interface generated by one or more graphics processors 1108.

[0106] In at least one embodiment, one or more processors 1102 each include one or more processor cores 1107 to process instructions which, when executed, implement the operations for system and user software. In at least one embodiment, each of the one or more processor cores 1107 is configured to process a specific instruction set 1109. In at least one embodiment, instruction set 1109 can facilitate Complex Instruction Set Computing (CISC), Reduced Instruction Set Computing (RISC), or computing via a Very Long Instruction Word (VLIW). In at least one embodiment, processor cores 1107 can each process a different instruction set 1109, which can include instructions to facilitate emulation of other instruction sets. In at least one embodiment, processor core 1107 can include other processing devices, such as a digital signal processor (DSP).

[0107] In at least one embodiment, processor 1102 includes cache memory 1104. In at least one embodiment, processor 1102 can have single-level or multi-level internal caches. In at least one embodiment, cache memory is shared among multiple components of processor 1102. In at least one embodiment, processor 1102 also uses an external cache, which can be shared by processor cores 1107, such as a level three (L3) cache or last level cache (LLC) (not shown), which can be shared among processor cores 1107 using known cache coherency techniques. In at least one embodiment, processor 1102 additionally includes a register file 1106 that can include different types of registers such as integer registers, floating point registers, status registers, and instruction pointer registers. In at least one embodiment, register file 1106 can include a general register file, a set of integer registers, or other registers.

[0108] In at least one embodiment, one or more processors 1102 are coupled with one or more interface buses 1110 for passing communication signals between processor 1102 and other components of system 1100. In at least one embodiment, one or more of interface buses 1110 can be versions of a Peripheral Component Interconnect (PCI) bus or PCI Express bus. In at least one embodiment, one or more of interface buses 1110 can be versions of an Industry Standard Architecture (ISA) bus, a MicroChannel Architecture (MCA) bus, an Intelll bus, a VESA local bus, an Accelerated Graphics Port (AGP) bus, a Video Electronics Standards Association (VESA) bus, or other suitable bus. In at least one embodiment, one or more of interface buses 1110 can be a bus using time-division multiplexing, a bus using Throttling, or another suitable bus structure at one or more embodiments. At least one embodiment of an interface bus 1110 can be operated at a voltage potential different from that of processor 1102, such as a 3.3 volt bus operated at a different voltage potential from a 1.8 volt processor 1102.

[0109] In at least one embodiment, memory device 1120 can be a Dynamic Random Access Memory (DRAM) device, a Static Random Access Memory (SRAM) device, a flash memory device, a phase- change memory device, or some other suitable memory device having a sufficient bandwidth to perform as a processor memory. In at least one embodiment, memory device 1120 can be used as system memory for processor system 1100, to store data 1122 and instructions 1121 for use when one or more processors 1102 executes an application or process. In at least one embodiment, memory controller 1116 also couples with an optional external graphics processor 1112, which can communicate with one or more graphics processors 1108 in processor 1102 to perform graphics and media operations. In at least one embodiment, a display device 1111 can be connected to processor 1102. In at least one embodiment, display device 1111 can include one or more of an internal display device, as in a mobile electronic device or a laptop device or an external display device attached via a display interface (such as DisplayPort, etc.). In at least one embodiment, display device 1111 can include a head mounted display (HMD) such as a stereoscopic display device for use in virtual reality (VR) applications or augmented reality (AR) applications.

[0110] In at least one embodiment, platform controller hub 1130 enables peripherals coupled to bridge 1122 to interact with a processor and / or each other over high-speed I / O buses 1120 and 1110. In at least one embodiment, I / O peripherals include, but are not limited to, an audio controller 1146, a network controller 1134, a firmware interface 1128, a wireless transceiver 1126, a touch sensor 1125, a data storage device 1124 (e.g., solid-state drive (SSD), floppy drive, optical drive, etc.). In at least one embodiment, data storage device 1124 can communicate via a storage interface (e.g., SATA) or via a peripheral bus, such as a Peripheral Component Interconnect bus (e.g., PCI, PCI Express). In at least one embodiment, touch sensor 1125 can include a touch screen sensor, a pressure sensor, or a fingerprint sensor. In at least one embodiment, wireless transceiver 1126 can be a Wi-Fi transceiver, a Bluetooth transceiver, or a mobile network transceiver such as a 3G, 4G, or Long Term Evolution (LTE) transceiver. In at least one embodiment, firmware interface 1128 enables communication with system firmware, and can be, for example, a unified extensible firmware interface (UEFI) or the like. In at least one embodiment, network controller 1134 can enable a network connection to a wired network. In at least one embodiment, a high-performance network controller (not shown) couples with interface bus 1110. In at least one embodiment, audio controller 1146 is a multi-channel high definition audio controller. In at least one embodiment, processing system 1100 includes an optional legacy I / O controller 1140 for coupling legacy (e.g., Personal System 2 (PS / 2)) devices to system 1100. In at least one embodiment, platform controller hub 1130 can also connect to one or more Universal Serial Bus (USB) controllers 1142 connect input devices, such as keyboard and mouse 1143 combinations, camera 1144, or other USB input devices.

[0111] In at least one embodiment, memory controller 1116 and platform controller hub 1130 can be integrated into a discrete external graphics processor, such as external graphics processor 1112. In at least one embodiment, platform controller hub 1130 and / or memory controller 1116 can be external to one or more processor(s) 1102. For example, in at least one embodiment, system 1100 can include an external memory controller 1116 and platform controller hub 1130, which can be configured as a memory controller hub and peripheral controller hub within a system chipset that is in communication with a processor(s) 1102.

[0112] Inference and / or training logic 715 are used to perform inferencing and / or training operations associated with one or more embodiments. Details regarding inference and / or training logic 715 are provided below in conjunction with FIGS. 7 and 8. Figure 7A and / or Figure 7BDetails regarding the inference and / or training logic 715 are provided. In at least one embodiment, some or all of inference and / or training logic 715 can be incorporated with graphics processor 1100. For example, in at least one embodiment, the training and / or inference techniques described herein can use one or more ALUs embodied in a graphics processor. Further, in at least one embodiment, the inference and / or training operations described herein can be accomplished with logic other than that shown. Figure 7A or Figure 7B In at least one embodiment, weight parameters can be stored in on-chip or off-chip memory and / or registers (shown or not) that configure ALUs of a graphics processor to perform one or more machine learning algorithms, neural network architectures, use cases, or training techniques described herein.

[0113] Such components can be used to train one or more generative models to be able to generate scene graphs for unlabeled data, as well as generate scenes based on those scene graphs that are indistinguishable from real scenes.

[0114] Figure 12 is a block diagram of a processor 1200 having one or more processor cores 1202A-1202N, an integrated memory controller 1214, and an integrated graphics processor 1208, according to at least one embodiment. In at least one embodiment, processor 1200 can include additional cores, up to and including an additional core 1202N denoted by a dashed line. In at least one embodiment, each processor core 1202A-1202N includes one or more internal cache units 1204A-1204N. In at least one embodiment, each processor core can also access one or more shared cache units 1206.

[0115] In at least one embodiment, internal cache units 1204A-1204N and shared cache unit 1206 represent a cache memory hierarchy within processor 1200. In at least one embodiment, cache memory units 1204A-1204N can include at least one level of instruction and data caches per processor core and one or more shared level 2 caches in a shared L2, level 3 (L3), level 4 (L4), or other level cache, where the highest level of cache prior to main memory is classified as an LLC. In at least one embodiment, cache coherence logic maintains coherence between various cache units 1206 and 1204A-1204N.

[0116] In at least one embodiment, processor 1200 also includes a set of one or more bus controller units 1216 and a system agent core 1210. In at least one embodiment, one or more bus controller units 1216 manage a set of peripheral buses, such as one or more PCI or PCIe buses. In at least one embodiment, system agent core 1210 provides management functionality for various processor components. In at least one embodiment, system agent core 1210 includes one or more integrated memory controllers 1214 to manage access to various external memory devices (not shown), including support for data bus protocols such as DDR SDRAM.

[0117] In at least one embodiment, one or more processor cores 1202A-1202N include support to run in multiple threads simultaneously. In at least one embodiment, system agent core 1210 includes components for coordination and operation of cores 1202A-1202N during multi-threaded processing. In at least one embodiment, system agent core 1210 can additionally include a power control unit (PCU), including logic and components to govern one or more power states of processor cores 1202A-1202N and graphics processor 1208.

[0118] In at least one embodiment, processor 1200 also includes graphics processor 1208, which can be configured to perform a graphics processing applications or a graphics processing- intensive applications. In at least one embodiment, graphics processor 1208 couples with shared cache unit 1206 and system agent core 1210, including one or more integrated memory controllers 1214. In at least one embodiment, system agent core 1210 also includes a display controller 1211 for driving one or more coupled displays to present graphics processor output to a

[0119] In at least one embodiment, ring based interconnect unit 1212 is used to couple the internal components of processor 1200. In at least one embodiment, an alternative interconnect unit can be used, such as a point-to-point interconnect, a switched interconnect, or other technology. In at least one embodiment, graphics processor 1208 couples with ring interconnect 1212 via I / O link 1213.

[0120] In at least one embodiment, I / O link 1213 represents at least one of a variety of I / O interconnects, including a package I / O interconnect that facilitates communication between various processor components and a high performance embedded memory module 1218 (e.g., an eDRAM module). In at least one embodiment, each of processor cores 1202A-1202N and graphics processor 1208 uses embedded memory module 1218 as a shared last level cache.

[0121] In at least one embodiment, processor cores 1202A-1202N are homogeneous cores executing a common instruction set architecture. In at least one embodiment, processor cores 1202A-1202N are heterogeneous with respect to instruction set architecture (ISA) in that one or more processor cores 1202A-1202N execute a common instruction set while one or more other processor cores 1202A-1202N execute a subset or a different instruction set. In at least one embodiment, processor cores 1202A-1202N are heterogeneous with respect to microarchitecture in that one or more cores have a relatively high power consumption while one or more other cores are power cores with lower power consumption. In at least one embodiment, processor 1200 can be implemented or realized as a SoC integrated circuit.

[0122] Inference and / or training logic 715 are used to perform inferencing and / or training operations associated with one or more embodiments. Details regarding inference and / or training logic 715 are provided below in conjunction with FIGS. 7A and / or 7B. In at least one embodiment, inference and / or training logic 715 can be used in graphics processor 1512, graphics cores 1202A-1202N, or other components illustrated in FIG. 15A, FIG. 15B, and / or FIG. 15C. Figure 7A and / or Figure 7B Details regarding inference and / or training logic 715 are provided below. In at least one embodiment, some or all of inference and / or training logic 715 can be incorporated into processor 1200. For example, in at least one embodiment, training and / or inferencing techniques described herein can use one or more ALUs embodied in graphics processor 1512, graphics cores 1202A-1202N, or other components illustrated in FIG. 15A, FIG. 15B, and / or FIG. 15C. Moreover, in at least one embodiment, inferencing and / or training operations described herein can be accomplished using logic other than that illustrated in FIG. 15A, FIG. 15B, and / or FIG. 15C. In at least one embodiment, weight parameters can be stored in on-chip or off-chip memory and / or registers (shown or not shown) that configure ALUs of graphics processor 1200 to perform one or more machine learning algorithms, neural network architectures, use cases, or training techniques described herein. Figure 12 Figure 7A or Figure 7B In at least one embodiment, weight parameters can be stored in on-chip or off-chip memory and / or registers (shown or not shown) that configure ALUs of graphics processor 1200 to perform one or more machine learning algorithms, neural network architectures, use cases, or training techniques described herein.

[0123] Such components can be used to train one or more generative models to be able to generate scene graphs for unlabeled data, as well as generate scenes based on those scene graphs that are indistinguishable from real scenes.

[0124] ​Virtualization computing platform

[0125] Figure 13 This is an example data flow diagram of process 1300 for generating and deploying an image processing and inference pipeline according to at least one embodiment. In at least one embodiment, process 1300 may be deployed for use with imaging devices, processing devices, and / or other device types at one or more facilities 1302. Process 1300 may be executed within training system 1304 and / or deployment system 1306. In at least one embodiment, training system 1304 may be used to train, deploy, and implement machine learning models (e.g., neural networks, object detection algorithms, computer vision algorithms, etc.) for use in deployment system 1306. In at least one embodiment, deployment system 1306 may be configured to offload processing and computing resources in a distributed computing environment to reduce the infrastructure requirements of facility 1302. In at least one embodiment, one or more applications in the pipeline may use or invoke services of deployment system 1306 (e.g., inference, visualization, computation, AI, etc.) during application execution.

[0126] In at least one embodiment, some applications used in the advanced processing and inference pipeline may use machine learning models or other AI to perform one or more processing steps. In at least one embodiment, a machine learning model may be trained at facility 1302 using data 1308 (e.g., imaging data) generated at facility 1302 (and stored on one or more Picture Archiving and Communication System (PACS) servers at facility 1302), imaging or sequencing data 1308 from another or more facilities, or a combination thereof. In at least one embodiment, training system 1304 may be used to provide applications, services, and / or other resources to generate a deployable machine learning model for the work of deploying system 1306.

[0127] In at least one embodiment, the model registry 1324 may be supported by an object storage system that supports version control and object metadata. In at least one embodiment, it may be available from within a cloud platform via, for example, cloud storage (e.g., Figure 14 The system uses a cloud-compatible application programming interface (API) to access object storage. In at least one embodiment, machine learning models within the model registry 1324 can be uploaded, listed, modified, or deleted by the developer or partner of the system interacting with the API. In at least one embodiment, the API can provide access to methods that allow users with appropriate credentials to associate models with applications, enabling the models to be executed as part of the containerized instantiation of the application.

[0128] In at least one embodiment, training pipeline 1404( Figure 14 ) can include scenarios in which a facility 1302 is training their own machine learning model, or has an existing machine learning model that needs to be optimized or updated. In at least one embodiment, imaging data 1308 generated by imaging devices, sequencing devices, and / or other types of devices can be received. In at least one embodiment, once imaging data 1308 is received, AI assisted annotation 1310 can be used to help generate annotations corresponding to imaging data 1308 to be used as ground truth data for a machine learning model. In at least one embodiment, AI assisted annotation 1310 can include one or more machine learning models (e.g., convolutional neural networks (CNNs)) that can be trained to generate annotations corresponding to certain types of imaging data 1308 (e.g., from certain devices). In at least one embodiment, AI assisted annotation 1310 can then be used directly, or can be adjusted or fine-tuned using annotation tools to generate ground truth data. In at least one embodiment, AI assisted annotation 1310, labeled clinical data 1312, or a combination thereof can be used as ground truth data to train a machine learning model. In at least one embodiment, a trained machine learning model can be referred to as an output model 1316, and can be used by deployment system 1306, as described herein.

[0129] In at least one embodiment, training pipeline 1404( Figure 14) can include situations in which facility 1302 needs a machine learning model for performing one or more processing tasks for one or more applications in deployment system 1306, but facility 1302 can not currently have such a machine learning model (or can not have a model that is optimized, efficient, or effective for this purpose). In at least one embodiment, an existing machine learning model can be selected from model registry 1324. In at least one embodiment, model registry 1324 can include machine learning models that are trained to perform a variety of different inferencing tasks on imaging data. In at least one embodiment, the machine learning models in model registry 1324 can have been trained on imaging data from different facilities (e.g., facilities located remotely from facility 1302). In at least one embodiment, a machine learning model can have been trained on imaging data from one location, two locations, or any number of locations. In at least one embodiment, when training on imaging data from a particular location, the training can be performed at that location, or at least in a manner that protects the confidentiality of the imaging data or limits the transfer of the imaging data offsite. In at least one embodiment, once a model is trained, or partially trained, at a location, the machine learning model can be added to model registry 1324. In at least one embodiment, the machine learning model can then be retrained or updated at any number of other facilities, and the retrained or updated model can be used in model registry 1324. In at least one embodiment, a machine learning model can then be selected from model registry 1324 (and referred to as output model 1316), and can be used in deployment system 1306 to perform one or more processing tasks for one or more applications of the deployment system.

[0130] In at least one embodiment, training pipeline 1404( Figure 14In at least one embodiment, scenario can include facility 1302 that requires a machine learning model for performing one or more processing tasks for deploying one or more applications in deployment system 1306, but facility 1302 can not currently have such a machine learning model (or can not have an optimized, efficient, or effective model). In at least one embodiment, due to population differences, robustness of training data used to train a machine learning model, diversity of training data anomalies, and / or other issues with training data, a machine learning model selected from model registry 1324 can not be fine-tuned or optimized for imaging data 1308 generated at facility 1302. In at least one embodiment, AI assisted annotation 1310 can be used to help generate annotations corresponding to imaging data 1308 for use as ground truth data to train or update a machine learning model. In at least one embodiment, labeled clinical data 1312 can be used as ground truth data to train a machine learning model. In at least one embodiment, retraining or updating a machine learning model can be referred to as model training 1314. In at least one embodiment, model training 1314 (e.g., AI assisted annotation 1310, labeled clinical data 1312, or a combination thereof) can be used as ground truth data to retrain or update a machine learning model. In at least one embodiment, a trained machine learning model can be referred to as output model 1316 and can be used by deployment system 1306, as described herein.

[0131] In at least one embodiment, deployment system 1306 can include software 1318, services 1320, hardware 1322, and / or other components, features, and functionality. In at least one embodiment, deployment system 1306 can include a software “stack” such that software 1318 can be built on top of services 1320, and can use services 1320 to perform some or all processing tasks, and services 1320 and software 1318 can be built on top of hardware 1322 and use hardware 1322 to perform processing, storage, and / or other computing tasks of deployment system. In at least one embodiment, software 1318 can include any number of different containers, where each container can execute an instantiation of an application. In at least one embodiment, each application can perform one or more processing tasks in a high-level processing and inference pipeline (e.g., inference, object detection, feature detection, segmentation, image enhancement, calibration, etc.). In at least one embodiment, a high-level processing and inference pipeline can be defined based on a selection of different containers desired or required to process imaging data 1308 (e.g., to convert output back to a usable data type, in addition to receiving and configuring containers for use by each container with imaging data for use and / or use by facility 1302 after processing through the pipeline. In at least one embodiment, a combination of containers within software 1318 (e.g., which make up a pipeline) can be referred to as a virtual instrument (as described in greater detail herein), and a virtual instrument can utilize services 1320 and hardware 1322 to perform some or all processing tasks of applications instantiated in containers.

[0132] In at least one embodiment, a data processing pipeline can receive input data (e.g., imaging data 1308) in a specific format in response to an inference request (e.g., a request from a user of deployment system 1306). In at least one embodiment, input data can represent one or more images, videos, and / or other data representations generated by one or more imaging devices. In at least one embodiment, data can be pre-processed as part of a data processing pipeline to prepare data for processing by one or more applications. In at least one embodiment, post-processing can be performed on output of one or more inference tasks or other processing tasks of a pipeline to prepare output data for a next application and / or to prepare output data for transmission and / or use by a user (e.g., in response to an inference request). In at least one embodiment, inference tasks can be performed by one or more machine learning models, such as trained or deployed neural networks, which can include output models 1316 of training system 1304.

[0133] In at least one embodiment, the tasks of the data processing pipeline can be encapsulated in containers, each container representing a discrete, fully functional instantiation of an application and a virtualized computing environment capable of referencing a machine learning model. In at least one embodiment, containers or applications can be published to a private (e.g., limited access) area of ​​a container registry (described in more detail herein), and trained or deployed models can be stored in a model registry 1324 and associated with one or more applications. In at least one embodiment, an image of an application (e.g., a container image) can be used in the container registry, and once a user selects an image from the container registry for deployment in the pipeline, that image can be used to generate containers for instantiation of the application for use by the user's system.

[0134] In at least one embodiment, a developer (e.g., a software developer, clinician, physician, etc.) can develop, publish, and store an application (e.g., as a container) for performing image processing and / or inference on provided data. In at least one embodiment, a software development kit (SDK) associated with the system can be used to perform development, publication, and / or storage (e.g., to ensure that the developed application and / or container conforms to or is compatible with the system). In at least one embodiment, the developed application can be tested locally using the SDK (e.g., at a first facility, testing data from a first facility), the SDK serving as a system (e.g.,...). Figure 14 System 1400 may support at least some services 1320. In at least one embodiment, since DICOM objects may contain one to hundreds of images or other data types, and due to variations in the data, the developer may be responsible for managing (e.g., setting up constructs for preprocessing built into the application, etc.) the extraction and preparation of incoming data. In at least one embodiment, once verified by system 1400 (e.g., for accuracy), the application becomes available in the container registry for user selection and / or implementation to perform one or more processing tasks on data at the user's facility (e.g., a second facility).

[0135] In at least one embodiment, the developer can then share the application or container over a network for the system (e.g., Figure 14of the system 1400) by users. In at least one embodiment, completed and validated applications or containers can be stored in a container registry, and related machine learning models can be stored in a model registry 1324. In at least one embodiment, a requesting entity (which provides an inference or image processing request) can browse the container registry and / or model registry 1324 to obtain applications, containers, datasets, machine learning models, etc., select a desired combination of elements to include in a data processing pipeline, and submit an image processing request. In at least one embodiment, a request can include input data necessary to perform the request (and, in some examples, data related to a patient), and / or can include a selection of applications and / or machine learning models to be executed in processing the request. In at least one embodiment, a request can then be passed to one or more components of the deployment system 1306 (e.g., a cloud) to perform processing of the data processing pipeline. In at least one embodiment, processing by the deployment system 1306 can include referencing elements (e.g., applications, containers, models, etc.) selected from the container registry and / or model registry 1324. In at least one embodiment, once results are generated through the pipeline, the results can be returned to a user for review (e.g., for review in a viewing application suite executed on a local, on-premises workstation or terminal).

[0136] In at least one embodiment, to help process or execute applications or containers in a pipeline, services 1320 can be utilized. In at least one embodiment, services 1320 can include computing services, artificial intelligence (Al) services, visualization services, and / or other service types. In at least one embodiment, services 1320 can provide functionality that is common to one or more applications in software 1318, and thus functionality can be abstracted as a service that can be called or utilized by applications. In at least one embodiment, functionality provided by services 1320 can run dynamically and more efficiently, while also allowing applications to process data in parallel (e.g., using Figure 14the parallel computing platform 1430) to scale well. In at least one embodiment, rather than requiring each application that requires the same functionality provided by a shared service 1320 to have a respective instance of the service 1320, the service 1320 can be shared among and between various applications. In at least one embodiment, as a non-limiting example, a service can include an inference server or engine that can be used to perform detection or segmentation tasks. In at least one embodiment, a model training service can be included, which can provide machine learning model training and / or retraining capabilities. In at least one embodiment, a data augmentation service can also be included, which can provide GPU-accelerated data (e.g., DICOM, RIS, CIS, REST-compliant, RPC, raw, etc.) extraction, resizing, expansion, and / or other augmentations. In at least one embodiment, a visualization service can be used, which can add image rendering effects (e.g., ray tracing, rasterization, de-noising, sharpening, etc.) to add realism to two-dimensional (2D) and / or three-dimensional (3D) models. In at least one embodiment, a virtual instrument service can be included, which provides beamforming, segmentation, inference, imaging, and / or support to other applications within a pipeline of a virtual instrument.

[0137] In at least one embodiment, where the services 1320 include an AI service (e.g., an inference service), as part of execution of an application, one or more machine learning models can be executed by invoking (e.g., as an API call) the inference service (e.g., inference server) to execute one or more machine learning models or processing thereof. In at least one embodiment, where another application includes one or more machine learning models for a segmentation task, the application can invoke the inference service to execute the machine learning model for performing one or more processing operations associated with the segmentation task. In at least one embodiment, software 1318 implementing an advanced processing and inference pipeline, which includes a segmentation application and an anomaly detection application, can be pipelined because each application can invoke the same inference service to perform one or more inference tasks. In at least one embodiment, hardware 1322 can include a GPU, CPU, graphics card, AI / deep learning system (e.g., an AI supercomputer such as NVIDIA’s DGX), a cloud platform, or a combination thereof.

[0138] In at least one embodiment, different types of hardware 1322 can be used to provide efficient, specially-built support for software 1318 and services 1320 in deployment system 1306. In at least one embodiment, GPU processing can be implemented for local processing within an AI / deep learning system, in a cloud system, and / or in other processing components of deployment system 1306 (e.g., at facility 1302) to improve efficiency, accuracy, and performance of image processing and generation. In at least one embodiment, software 1318 and / or services 1320 can be optimized for GPU processing, by way of non-limiting example with respect to deep learning, machine learning, and / or high performance computing. In at least one embodiment, at least some of computing environments of deployment system 1306 and / or training system 1304 can be executed in a data center, one or more supercomputer or high performance computer systems with GPU-optimized software (e.g., a combination of hardware and software of NVIDIA DGX systems). In at least one embodiment, hardware 1322 can include any number of GPUs that can be called upon to perform data processing in parallel, as described herein. In at least one embodiment, a cloud platform can also include GPU processing for GPU-optimized execution of deep learning tasks, machine learning tasks, or other computing tasks. In at least one embodiment, an AI / deep learning supercomputer and / or GPU-optimized software (e.g., as provided on NVIDIA’s DGX systems) can be used as a hardware abstraction and extension platform to execute a cloud platform (e.g., NVIDIA’s NGC) in at least one embodiment. In at least one embodiment, a cloud platform can integrate an application container clustering system or orchestration system (e.g., KUBERNETES) across multiple GPUs to enable seamless scaling and load balancing.

[0139] Figure 14 is a system diagram of an example system 1400 for generating and deploying imaging deployment pipelines, in accordance with at least one embodiment. In at least one embodiment, system 1400 can be used to implement processes 1300 and / or other processes of FIG. 13, including advanced processing and inference pipelines. In at least one embodiment, system 1400 can include training system 1304 and deployment system 1306. In at least one embodiment, training system 1304 and deployment system 1306 can be implemented using software 1318, services 1320, and / or hardware 1322, as described herein. Figure 13

[0140] ​In at least one embodiment, system 1400 (e.g., training system 1304 and / or deployment system 1306) can be implemented in a cloud computing environment (e.g., using cloud 1426). In at least one embodiment, system 1400 can be implemented locally (with respect to a medical service facility), or as a combination of cloud computing resources and local computing resources. In at least one embodiment, access to APIs in cloud 1426 can be limited to authorized users by instituting security measures or protocols. In at least one embodiment, security protocols can include network tokens that can be signed by an authentication (e.g., AuthN, AuthZ, Gluecon, etc.) service and can carry appropriate authorization. In at least one embodiment, APIs (described herein) of a virtual instrument or other instances of system 1400 can be limited to a set of public IPs that have been vetted or authorized for interaction.

[0141] In at least one embodiment, various components of system 1400 can communicate information between each other using any of a plurality of different network types, including but not limited to local area networks (LANs) and / or wide area networks (WANs) via wired and / or wireless communication protocols. In at least one embodiment, communication between facilities and components of system 1400 (e.g., for sending inference requests, for receiving results of inference requests, etc.) can be communicated through one or more data buses, wireless data protocols (Wi-Fi), wired data protocols (e.g., Ethernet), etc.

[0142] In at least one embodiment, similar to training pipelines 1302 described herein with respect to Figure 13 In at least one embodiment, training system 1304 can execute training pipeline 1404. In at least one embodiment, where deployment system 1306 is to use one or more machine learning models in deployment pipeline 1410, training pipeline 1404 can be used to train or retrain one or more (e.g., pre-trained) models, and / or implement one or more pre-trained models 1406 (e.g., without retraining or updating). In at least one embodiment, as a result of training pipeline 1404, an output model 1316 can be generated. In at least one embodiment, training pipeline 1404 can include any number of processing steps, such as but not limited to conversion or adaptation of imaging data (or other input data). In at least one embodiment, different training pipelines 1404 can be used for different machine learning models used by deployment system 1306. In at least one embodiment, similar to training pipeline 1302 described herein with respect to Figure 13 Training pipeline 1404 of a first example described herein with respect to Figure 13 Training pipeline 1404 of a second example described herein with respect to Figure 13The training pipeline 1404 of the third example described can be used for the third machine learning model. In at least one embodiment, any combination of tasks within the training system 1304 can be used depending on the requirements of each respective machine learning model. In at least one embodiment, one or more machine learning models can already be trained and ready for deployment, so the training system 1304 can not perform any processing on the machine learning model and the one or more machine learning models can be implemented by the deployment system 1306.

[0143] In at least one embodiment, the output model 1316 and / or the pre-trained model 1406 can include any type of machine learning model, depending on implementation or embodiment. In at least one embodiment and without limitation thereto, machine learning models used by the system 1400 can include using linear regression, logistic regression, decision trees, support vector machines (SVM), Naive Bayes, k- nearest neighbors (Knn), k-means clustering, random forests, dimensionality reduction algorithms, gradient boosting algorithms, neural networks (e.g., autoencoders, convolutional, recurrent, perceptrons, long / short-term memory (LSTM), Hopfield, Boltzmann, deep belief, deconvolutional, generative adversarial, liquid state machines, etc.), and / or other types of machine learning models.

[0144] In at least one embodiment, the training pipeline 1404 can include AI-assisted annotation, as described herein with respect to at least Figure 15BIn at least one embodiment, labeled clinical data 1312 (e.g., traditional annotations) can be generated by any number of techniques. In at least one embodiment, annotations or other labels can be generated in a drawing program (e.g., annotation program), a computer aided design (CAD) program, a labeling program, another type of application suitable for generating annotations or labels for ground truth, and / or can be hand drawn, in some examples. In at least one embodiment, ground truth data can be synthetically generated (e.g., from computer models or renderings), realistically generated (e.g., designed and generated from real world data), automatically generated by a machine (e.g., using feature analysis and learning to extract features from data and then generate labels), manually annotated (e.g., by a labeler or annotation specialist defining locations of labels), and / or combinations thereof. In at least one embodiment, for each instance of imaging data 1308 (or other data types used by a machine learning model), there can be corresponding ground truth data generated by training system 1304. In at least one embodiment, AI-assisted annotation can be performed as part of deployment pipeline 1410; in addition to or instead of AI-assisted annotation included in training pipeline 1404. In at least one embodiment, system 1400 can include a multi-tiered platform that can include a software tier of diagnostic applications (or other application types) (e.g., software 1318) that can perform one or more medical imaging and diagnostic functions. In at least one embodiment, system 1400 can be communicatively coupled to (e.g., via encrypted links) a PACS server network of one or more facilities. In at least one embodiment, system 1400 can be configured to access and reference data from a PACS server to perform operations such as training machine learning models, deploying machine learning models, image processing, inferencing, and / or other operations.

[0145] In at least one embodiment, a software tier can be implemented as a secure, encrypted, and / or authenticated API through which applications or containers can be invoked (e.g., called) from an external environment (e.g., facility 1302). In at least one embodiment, an application can then call or execute one or more services 1320 to perform computing, AI, or visualization tasks associated with respective applications, and software 1318 and / or services 1320 can utilize hardware 1322 to perform processing tasks in an efficient and effective manner.

[0146] In at least one embodiment, deployment system 1306 can execute deployment pipelines 1410. In at least one embodiment, deployment pipelines 1410 can include any number of applications that can be sequential, non-sequential, or otherwise applied to imaging data (and / or other data types) - including AI assisted annotation - generated by imaging devices, sequencing devices, genomics devices, etc., as described above. In at least one embodiment, a deployment pipeline 1410 for an individual device can be referred to as a virtual instrument for a device (e.g., a virtual ultrasound instrument, a virtual CT scan instrument, a virtual sequencing instrument, etc.), as described herein. In at least one embodiment, there can be more than one deployment pipeline 1410 for a single device, depending on information desired from data generated by a device. In at least one embodiment, where anomalies are desired to be detected from MRI machines, there can be a first deployment pipeline 1410, and where image enhancement is desired from output of MRI machines, there can be a second deployment pipeline 1410.

[0147] In at least one embodiment, image generation applications can include processing tasks that include use of machine learning models. In at least one embodiment, a user can wish to use their own machine learning model, or select a machine learning model from model registry 1324. In at least one embodiment, a user can implement their own machine learning model or select a machine learning model for inclusion in an application that performs a processing task. In at least one embodiment, applications can be selectable and customizable, and by defining a construction of an application, deployment and implementation of an application for a particular user is presented as a more seamless user experience. In at least one embodiment, by leveraging other features of system 1400 (e.g., services 1320 and hardware 1322), deployment pipelines 1410 can be more user friendly, provide easier integration, and produce more accurate, efficient, and timely results.

[0148] In at least one embodiment, deployment system 1306 can include a user interface 1414 (e.g., a graphical user interface, a web interface, etc.) that can be used to select applications to include in deployment pipelines 1410, arrange applications, modify or change applications or parameters or constructions thereof, use and interact with deployment pipelines 1410 during setup and / or deployment, and / or otherwise interact with deployment system 1306. In at least one embodiment, although not shown with respect to training system 1304, user interface 1414 (or a different user interface) can be used to select models for use in deployment system 1306, for selecting models for training or retraining in training system 1304, and / or for otherwise interacting with training system 1304.

[0149] In at least one embodiment, in addition to application orchestration system 1428, pipeline manager 1412 can be used to manage interactions between applications or containers of deployment pipeline 1410 and services 1320 and / or hardware 1322. In at least one embodiment, pipeline manager 1412 can be configured to facilitate interactions from application to application, from application to service 1320, and / or from application or service to hardware 1322. In at least one embodiment, although illustrated as included in software 1318, this is not intended to be limiting, and in some examples (e.g., as shown by cc), pipeline manager 1412 can be included in services 1320. In at least one embodiment, application orchestration system 1428 (e.g., Kubernetes, DOCKER, etc.) can include a container orchestration system that can group applications into containers as logical units for orchestration, management, scaling, and deployment. In at least one embodiment, by associating applications (e.g., rebuilt applications, split applications, etc.) from deployment pipeline 1410 with individual containers, each application can execute in a self-contained environment (e.g., at kernel level) to improve speed and efficiency. Figure 12

[0150] ​In at least one embodiment, each application and / or container (or image thereof) can be separately developed, modified, and deployed (e.g., a first user or developer can develop, modify, and deploy a first application, a second user or developer can develop, modify, and deploy a second application separate from first user or developer), which can allow for focus and attention to tasks of a single application and / or container without being impeded by tasks of another application or container. In at least one embodiment, pipeline manager 1412 and application coordination system 1428 can facilitate communication and collaboration between different containers or applications. In at least one embodiment, as long as intended inputs and / or outputs of each container or application are known to system (e.g., based on construction of application or container), application coordination system 1428 and / or pipeline manager 1412 can facilitate communication between and among each application or container and sharing of resources. In at least one embodiment, as one or more applications or containers in deployment pipeline 1410 can share same services and resources, application coordination system 1428 can coordinate, load balance, and determine sharing of services or resources between and among various applications or containers. In at least one embodiment, a scheduler can be used to track resource needs of applications or containers, current or planned use of these resources, and resource availability. Accordingly, in at least one embodiment, a scheduler can allocate resources to different applications and among and between applications, taking into account needs and availability of system. In some examples, a scheduler (and / or other components of application coordination system 1428) can determine resource availability and distribution based on constraints imposed on system (e.g., user constraints), such as quality of service (QoS), urgency of data output (e.g., to determine whether to perform real-time processing or delayed processing), etc.

[0151] In at least one embodiment, services 1320 utilized by and shared by applications or containers in deployment system 1306 can include compute services 1416, AI services 1418, visualization services 1420, and / or other service types. In at least one embodiment, applications can invoke (e.g., execute) one or more services 1320 to perform processing operations for applications. In at least one embodiment, applications can utilize compute services 1416 to perform supercomputing or other high performance computing (HPC) tasks. In at least one embodiment, one or more compute services 1416 can be utilized to perform parallel processing (e.g., using parallel computing platform 1430) to process data substantially simultaneously by one or more applications and / or one or more tasks of a single application. In at least one embodiment, parallel computing platform 1430 (e.g., NVIDIA’s CUDA) can implement general purpose computing on GPUs (GPGPU) (e.g., GPU 1422). In at least one embodiment, software layers of parallel computing platform 1430 can provide access to virtual instruction sets and parallel computing elements of GPUs to execute compute kernels. In at least one embodiment, parallel computing platform 1430 can include memory, and in some embodiments, memory can be shared between and among multiple containers, and / or between and among different processing tasks within a single container. In at least one embodiment, inter-process communication (IPC) calls can be generated for multiple containers and / or multiple processes within a container to use same data from a shared memory segment of parallel computing platform 1430 (e.g., where multiple different stages of an application or applications are processing same information). In at least one embodiment, rather than copying data and moving data to different locations in memory (e.g., read / write operations), same data in same location in memory can be used for any number of processing tasks (e.g., at same time, at different times, etc.). In at least one embodiment, as data is used to generate new data as a result of processing, this information of new location of data can be stored and shared between various applications. In at least one embodiment, location of data, as well as location of updated or modified data, can be part of a definition of how to understand a payload in a container.

[0152] In at least one embodiment, AI services 1418 can be utilized to perform inferencing services for executing machine learning models associated with applications (e.g., task is to perform one or more processing tasks for an application). In at least one embodiment, AI services 1418 can utilize AI system 1424 to execute machine learning models (e.g., neural networks such as CNNs) for segmentation, reconstruction, object detection, feature detection, classification, and / or other inferencing tasks. In at least one embodiment, an application of deployment pipeline 1410 can use one or more output models 1316 from training system 1304 and / or other models of an application to perform inferencing on imaging data. In at least one embodiment, two or more examples of inferencing can be available using application coordination system 1428 (e.g., a scheduler). In at least one embodiment, a first category can include a high priority / low latency path, which can implement a higher service level agreement, such as for performing inferencing on urgent requests in emergency situations, or for radiologists during a diagnosis process. In at least one embodiment, a second category can include a standard priority path, which can be used for requests that can not be urgent or can have analysis performed at a later time. In at least one embodiment, application coordination system 1428 can allocate resources (e.g., services 1320 and / or hardware 1322) based on a priority path for different inferencing tasks of AI services 1418.

[0153] In at least one embodiment, shared storage can be installed to AI services 1418 in system 1400. In at least one embodiment, shared storage can operate as a cache (or other storage device type) and can be used to process inference requests from applications. In at least one embodiment, when an inference request is submitted, a set of API instances of deployment system 1306 can receive the request and can select one or more instances (e.g., for best fit, for load balancing, etc.) to process the request. In at least one embodiment, to process the request, the request can be input into a database, a machine learning model can be located from model registry 1324 if not already in cache, a validation step can ensure that appropriate machine learning model is loaded into cache (e.g., shared storage), and / or a copy of the model can be saved to cache. In at least one embodiment, if an application has not already been running or there are not enough instances of an application, a scheduler (e.g., of pipeline manager 1412) can be used to start the application referenced in the request. In at least one embodiment, if an inference server has not already been started to execute the model, an inference server can be started. Each model can start any number of inference servers. In at least one embodiment, in a pull model of clustering inference servers, a model can be cached whenever load balancing is favorable. In at least one embodiment, inference servers can be statically loaded into respective distributed servers.

[0154] In at least one embodiment, inference can be performed using inference servers running in containers. In at least one embodiment, an instance of an inference server can be associated with a model (and optionally multiple versions of a model). In at least one embodiment, if an instance of an inference server does not exist when a request to perform inference on a model is received, a new instance can be loaded. In at least one embodiment, when an inference server is started, a model can be passed to the inference server so that the same container can be used to service different models as long as the inference server is run as a different instance.

[0155] In at least one embodiment, during application execution, an inference request for a given application can be received and a container (e.g., an instance hosting an inference server) can be loaded (if not already loaded) and a launcher can be invoked. In at least one embodiment, pre-processing logic in a container can load, decode, and / or perform any additional pre-processing on incoming data (e.g., using CPU and / or GPU). In at least one embodiment, once data is ready for inference, a container can infer on data as needed. In at least one embodiment, this can include a single inference call on one image (e.g., a hand X-ray), or can require inference on hundreds of images (e.g., a chest CT). In at least one embodiment, an application can summarize results before completion, which can include, without limitation, a single confidence score, a pixel-level segmentation, a voxel-level segmentation, generating a visualization, or generating text to summarize results. In at least one embodiment, different priorities can be assigned for different models or applications. For example, some models can have real-time (TAT less than 1 minute) priority, while other models can have lower priority (e.g., TAT less than 10 minutes). In at least one embodiment, model execution time can be measured from a requesting authority or entity, and can include cooperative network traversal time as well as execution time of an inference service.

[0156] In at least one embodiment, transfer of requests between service 1320 and inference applications can be hidden behind a software development kit (SDK), and robust transfer can be provided through queues. In at least one embodiment, requests will be placed in queues through an API for individual application / tenant ID combinations, and SDK will pull requests from queues and provide requests to applications. In at least one embodiment, a name of a queue can be provided in an environment from which SDK will pick up queues. In at least one embodiment, asynchronous communication through queues can be useful because it can allow any instance of an application to pick up work when it is available. Results can be transferred back through queues to ensure no data loss. In at least one embodiment, queues can also provide an ability to split work, because highest priority work can go into a queue that connects to most instances of an application, while lowest priority work can go into a queue that connects to a single instance that processes tasks in order of receipt. In at least one embodiment, an application can run on GPU-accelerated instances that are spawned in cloud 1426, and inference service can perform inference on GPUs.

[0157] In at least one embodiment, visualization service 1420 can be utilized to generate visualizations for viewing application and / or deployment pipeline 1410 output. In at least one embodiment, visualization service 1420 can utilize GPU 1422 to generate visualizations. In at least one embodiment, visualization service 1420 can implement rendering effects such as ray tracing to generate higher quality visualizations. In at least one embodiment, visualizations can include, without limitation, 2D image rendering, 3D volume rendering, 3D volume reconstruction, 2D tomographic slices, virtual reality displays, augmented reality displays, etc. In at least one embodiment, virtualization environment can be used to generate virtual interactive displays or environments (e.g., virtual environments) for interaction by system users (e.g., doctors, nurses, radiologists, etc.). In at least one embodiment, visualization service 1420 can include internal visualizers, movie and / or other rendering or image processing capabilities or functionality (e.g., ray tracing, rasterization, internal optics, etc.).

[0158] In at least one embodiment, hardware 1322 can include GPU 1422, AI system 1424, cloud 1426, and / or any other hardware used to execute training system 1304 and / or deployment system 1306. In at least one embodiment, GPU 1422 (e.g., NVIDIA’s TESLA and / or QUADRO GPUs) can include any number of GPUs that can be used to perform processing tasks for any features or functionality of computing service 1416, AI service 1418, visualization service 1420, other services, and / or software 1318. For example, for AI service 1418, GPU 1422 can be used to perform pre-processing on imaging data (or other data types used by machine learning models), post-processing on outputs of machine learning models, and / or perform inferencing (e.g., to execute machine learning models). In at least one embodiment, cloud 1426, AI system 1424, and / or other components of system 1400 can use GPU 1422. In at least one embodiment, cloud 1426 can include a GPU-optimized platform for deep learning tasks. In at least one embodiment, AI system 1424 can use GPUs, and one or more AI systems 1424 can be used to perform cloud 1426 (or at least portions of tasks that are deep learning or inferencing). Likewise, although hardware 1322 is shown as discrete components, this is not intended to be limiting, and any component of hardware 1322 can be combined with, or utilized by, any other component of hardware 1322.

[0159] In at least one embodiment, AI system 1424 can include a purpose-built computing system (e.g., a supercomputer or HPC) configured for inferencing, deep learning, machine learning, and / or other artificial intelligence tasks. In at least one embodiment, AI system 1424 (e.g., NVIDIA’s DGX) can include software (e.g., a software stack) that can use multiple GPUs 1422 to perform split-GPU optimizations in addition to CPUs, RAM, storage, and / or other components, features, or functions. In at least one embodiment, one or more AI systems 1424 can be implemented in cloud 1426 (e.g., in a data center) to perform some or all of AI-based processing tasks of system 1400.

[0160] In at least one embodiment, cloud 1426 can include a GPU-accelerated infrastructure (e.g., NVIDIA’s NGC) that can provide a GPU-optimized platform for performing processing tasks of system 1400. In at least one embodiment, cloud 1426 can include AI system 1424 for performing one or more AI-based tasks of system 1400 (e.g., as a hardware abstraction and scaling platform). In at least one embodiment, cloud 1426 can integrate with application coordination system 1428 that utilizes multiple GPUs to enable seamless scaling and load balancing between and among applications and services 1320. In at least one embodiment, cloud 1426 can be responsible for performing at least some services 1320 of system 1400, including compute services 1416, AI services 1418, and / or visualization services 1420, as described herein. In at least one embodiment, cloud 1426 can perform batched inferencing (e.g., performing NVIDIA’s TENSORRT), provide an accelerated parallel computing API and platform 1430 (e.g., NVIDIA’s CUDA), perform application coordination system 1428 (e.g., KUBERNETES), provide a graphics rendering API and platform (e.g., for ray tracing, 2D graphics, 3D graphics, and / or other rendering techniques to produce higher quality cinematic effects), and / or can provide other functionality for system 1400.

[0161] Figure 15A A dataflow graph for process 1500 for training, retraining, or updating a machine learning model is shown, in accordance with at least one embodiment. In at least one embodiment, process 1500 can be performed using, as a non-limiting example, NVIDIA’s Figure 14The process 1500 can be performed by the system 1400. In at least one embodiment, the process 1500 can utilize the services 1320 and / or hardware 1322 of the system 1400, as described herein. In at least one embodiment, the refined model 1512 generated by the process 1500 can be executed by the deployment system 1306 for one or more containerized applications in the deployment pipeline 1410.

[0162] In at least one embodiment, the model training 1314 can include retraining or updating the initial model 1504 (e.g., a pre-trained model) using new training data (e.g., new input data (such as the customer dataset 1506), and / or new ground truth data associated with the input data). In at least one embodiment, to retrain or update the initial model 1504, the output or loss layers of the initial model 1504 can be reset or deleted, and / or replaced with updated or new output or loss layers. In at least one embodiment, the initial model 1504 can have previously fine-tuned parameters (e.g., weights and / or biases) that are retained from previous training, so the training or retraining 1314 can not take as long or require as much processing as training a model from scratch. In at least one embodiment, during the model training 1314, by resetting or replacing the output or loss layers of the initial model 1504, the parameters of the new dataset can be updated and re-tuned as predictions are generated on the new customer dataset 1506 (e.g., image data 1308) based on loss calculations associated with the accuracy of the output or loss layers. Figure 13 In at least one embodiment, the model training 1314 can include retraining or updating the initial model 1504 (e.g., a pre-trained model) using new training data (e.g., new input data (such as the customer dataset 1506), and / or new ground truth data associated with the input data). In at least one embodiment, to retrain or update the initial model 1504, the output or loss layers of the initial model 1504 can be reset or deleted, and / or replaced with updated or new output or loss layers. In at least one embodiment, the initial model 1504 can have previously fine-tuned parameters (e.g., weights and / or biases) that are retained from previous training, so the training or retraining 1314 can not take as long or require as much processing as training a model from scratch. In at least one embodiment, during the model training 1314, by resetting or replacing the output or loss layers of the initial model 1504, the parameters of the new dataset can be updated and re-tuned as predictions are generated on the new customer dataset 1506 (e.g., image data 1308) based on loss calculations associated with the accuracy of the output or loss layers.

[0163] In at least one embodiment, the pre-trained model 1406 can be stored in a data store or registry (e.g., the data store 1302) for use by the deployment system 1306 to deploy the model 1406 to one or more containerized applications in the deployment pipeline 1410. Figure 13model registry 1324). In at least one embodiment, pre-trained models 1406 can have been trained, at least in part, at one or more facilities other than the facility executing process 1500. In at least one embodiment, to protect the privacy and rights of patients, subjects, or customers of different facilities, pre-trained models 1406 can have been trained locally using locally generated customer or patient data. In at least one embodiment, pre-trained models 1406 can be trained using cloud 1426 and / or other hardware 1322, but confidential, privacy protected patient data can not be transferred to, used by, or accessed by any component of cloud 1426 (or other non-local hardware). In at least one embodiment, if pre-trained models 1406 are trained using patient data from more than one facility, pre-trained models 1406 can have been individually trained for each facility before training on patient or customer data from another facility. In at least one embodiment, customer or patient data from any number of facilities can be used to train pre-trained models 1406 locally and / or externally, such as in a data center or other cloud computing infrastructure, for example, in cases where customer or patient data has been de-identified (e.g., by waiver, for experimental use, etc.), or where customer or patient data is included in a public dataset.

[0164] In at least one embodiment, when selecting an application to use in deployment pipeline 1410, a user can also select a machine learning model for use with the particular application. In at least one embodiment, a user can not have a model to use, so the user can select a pre-trained model 1406 to use with the application. In at least one embodiment, pre-trained models 1406 can not be optimized for generating accurate results on a customer dataset 1506 of a user’s facility (e.g., based on patient diversity, demographics, types of medical imaging devices used, etc.). In at least one embodiment, pre-trained models 1406 can be updated, retrained, and / or fine-tuned for use at individual facilities before being deployed into deployment pipeline 1410 for use with one or more applications.

[0165] In at least one embodiment, a user can select a pre-trained model 1406 to update, retrain, and / or fine-tune, and the pre-trained model 1406 can be referred to as an initial model 1504 for training system 1304 in process 1500. In at least one embodiment, a customer dataset 1506 (e.g., imaging data, genomic data, sequencing data, or other data types generated by equipment at a facility) can be used to perform model training 1314 (which can include, without limitation, transfer learning) on initial model 1504 to generate a refined model 1512. In at least one embodiment, ground truth data corresponding to customer dataset 1506 can be generated by training system 1304. In at least one embodiment, ground truth data can be generated at least in part by a clinician, scientist, physician, practitioner at a facility (e.g., as labeled clinical data 1312 in FIG. 13B). Figure 13

[0166] In at least one embodiment, AI-assisted annotation 1310 can be used in some examples to generate ground truth data. In at least one embodiment, AI-assisted annotation 1310 (e.g., implemented using an AI-assisted annotation SDK) can utilize a machine learning model (e.g., a neural network) to generate suggested or predicted ground truth data for a customer dataset. In at least one embodiment, a user 1510 can use annotation tools within a user interface (graphical user interface (GUI)) on computing device 1508.

[0167] In at least one embodiment, user 1510 can interact with GUI via computing device 1508 to edit or fine-tune annotations or automated annotations. In at least one embodiment, a polygon editing feature can be used to move vertices of a polygon to more precise or fine-tuned locations.

[0168] In at least one embodiment, once customer dataset 1506 has associated ground truth data, the ground truth data (e.g., from AI-assisted annotation, manual labeling, etc.) can be used during model training 1314 to generate refined model 1512. In at least one embodiment, customer dataset 1506 can be applied to initial model 1504 any number of times, and ground truth data can be used to update parameters of initial model 1504 until an acceptable level of accuracy is reached for refined model 1512. In at least one embodiment, once refined model 1512 is generated, refined model 1512 can be deployed within one or more deployment pipelines 1410 at a facility for performing one or more processing tasks with respect to medical imaging data.

[0169] ​In at least one embodiment, a refined model 1512 can be uploaded to pre-trained models 1406 in model registry 1324 for selection by another facility. In at least one embodiment, his process can be completed at any number of facilities such that refined model 1512 can be further refined any number of times on new data sets to generate more general purpose models.

[0170] Figure 15B is an example illustration of a client-server architecture 1532 for augmenting annotation tools with pre-trained annotation models, in accordance with at least one embodiment. In at least one embodiment, an AI-assisted annotation tool 1536 can be instantiated based on client-server architecture 1532. In at least one embodiment, annotation tools 1536 in an imaging application can assist radiologists, for example, in identifying organs and abnormalities. In at least one embodiment, an imaging application can include software tools that help a user 1510 identify a few extreme points on a particular organ of interest in a raw image 1534 (e.g., in a 3D MRI or CT scan), for example, and receive automatic annotation results for all 2D slices of a particular organ. In at least one embodiment, results can be stored as training data 1538 in a data store and used as ground truth data for training, for example, but not by way of limitation. In at least one embodiment, when a computing device 1508 sends extreme points for AI-assisted annotation 1310, a deep learning model, for example, can receive that data as input and return an inference result that segments an organ or abnormality. In at least one embodiment, a pre-instantiated annotation tool (e.g., AI-assisted annotation tool 1536B in Figure 15B AI-assisted annotation tools 1536B) can be augmented by making API calls (e.g., API call 1544) to a server, such as an annotation helper server 1540, which can include a set of pre-trained models 1542 stored in an annotation model registry, for example. In at least one embodiment, an annotation model registry can store pre-trained models 1542 (e.g., machine learning models, such as deep learning models) that are pre-trained to perform AI-assisted annotation on a particular organ or abnormality. In at least one embodiment, these models can be further updated by using a training pipeline 1404. In at least one embodiment, as new labeled clinical data 1312 is added, a pre-installed annotation tool can be improved over time.

[0171] Such components can be used to train one or more generative models to be able to generate scene graphs for unlabeled data, as well as generate scenes based on those scene graphs that are indistinguishable from real scenes.

[0172] Other variations are within the spirit of the present disclosure. Thus, while the disclosed technology is susceptible to various modifications and alternative constructions, certain illustrated embodiments thereof are shown in the drawings and have been described above in detail. It should, however, be understood that there is no intention to limit the disclosure to the specific form or forms disclosed, but on the contrary, the intention is to cover all modifications, alternative constructions, and equivalents falling within the spirit and scope of the disclosure, as defined in the appended claims.

[0173] Unless otherwise stated or contradicted by context, the use of the terms "a" and "an" and "the" and similar referents in the context of describing the disclosed embodiments (especially in the context of the following claims) are to be construed to cover both the singular and the plural, unless otherwise indicated by context. The terms "comprising," "having," "including," and "containing" are to be construed as open-ended terms (meaning "including, but not limited to") unless otherwise noted by context. The term "connected" (when used without modification) is to be construed as partly or wholly encompassed in, attached to, or joined together with, even if there are some intervening matters. Unless otherwise indicated herein, the reference herein to a range of values is intended merely as a shorthand method of referring individually to each separate value falling within the range, and each separate value is incorporated in the specification as if it were individually recited herein. Unless otherwise indicated or contradicted by context, the use of the term "set" (e.g., "set of items") or "subset" is to be construed as a non-empty set of one or more members. Furthermore, unless otherwise indicated or contradicted by context, the term "subset" of a corresponding set does not necessarily denote a proper subset of the corresponding set, but rather the subset and the corresponding set can be equal.

[0174] Unless explicitly indicated otherwise or contradicted by context, conjunction language such as phrases in the form "at least one of A, B, and C" or "at least one of A, B, and C" is to be construed in context as generally used to indicate items, clauses, etc., that can be A or B or C, or any non-empty subset of the set of A and B and C. For example, in the illustrative example of a set having three members, the conjunctive phrases "at least one of A, B, and C" and "at least one of A, B, and C" refer to any of the following sets: {A}, {B}, {C}, {A, B}, {A, C}, {B, C}, {A, B, C}. Thus, such conjunctive language is not generally intended to imply that certain embodiments require the existence of at least one of A, at least one of B, and at least one of C. Additionally, unless otherwise indicated or contradicted by context, the term "plurality" denotes a state of plurality (e.g., "a plurality of items" denotes multiple items). The number of items in a plurality of items is at least two, but can be more if explicitly indicated or indicated by context. Furthermore, unless otherwise indicated or clear from context, the phrase "based on" means "based at least in part on" rather than "based only on."

[0175] Unless otherwise indicated herein, or otherwise clearly contradicted by context, the operations of a process described herein can be performed in any suitable order. In at least one embodiment, processes such as those described herein (or variations and / or combinations thereof) are performed under the control of one or more computer systems configured with executable instructions to perform the operations of the processes, and are implemented as code (e.g., executable instructions, one or more computer programs or one or more applications) executing collectively on one or more processing units, by hardware or combinations thereof. In at least one embodiment, the code is stored on a computer-readable storage medium, example forms of which include non-transitory computer- readable storage media and tangible computer-readable storage media. In at least one embodiment, the code is stored on a computer-readable storage medium that excludes transitory signals (e.g., a propagating transient electric or electromagnetic transmission). In at least one embodiment, a computer-readable storage medium is a non-transitory computer-readable storage medium that excludes transitory signals (e.g., a propagating transient electric or electromagnetic transmission), but includes non-transitory data storage circuitry (e.g., buffers, cache, and queues). In at least one embodiment, the code (e.g., executable instructions or source code) is stored on a set of one or more non-transitory computer-readable storage media (or other memory for storing executable instructions) having stored thereon executable instructions that, as a result of being executed by one or more processors of a computer system (i.e., as a result of being executed), cause the computer system to perform operations described herein. In at least one embodiment, a set of non-transitory computer-readable storage media includes multiple non-transitory computer-readable storage media, and one or more of the individual non-transitory storage media of the multiple non-transitory computer-readable storage media lack all of the code, with the multiple non-transitory computer-readable storage media collectively storing the entire code. In at least one embodiment, executable instructions are executed by different processors, e.g., the non-transitory computer-readable storage media stores instructions and a main central processing unit (“CPU”) executes some instructions, while a graphics processing unit (“GPU”) executes other instructions. In at least one embodiment, different components of a computer system have separate processors and different processors execute different subsets of the instructions.

[0176] Accordingly, in at least one embodiment, a computer system is configured to implement one or more services that individually or collectively perform operations of processes described herein, and such a computer system is configured with applicable hardware and / or software to enable implementation of the operations. Moreover, a computer system implementing at least one embodiment of the present disclosure is a single device, and in another embodiment is a distributed computer system including multiple devices operating in different manners such that the distributed computer system performs operations described herein, and such that a single device does not perform all of the operations.

[0177] The use of any and all examples, or exemplary language (e.g., "such as") provided herein, is intended merely to better illuminate embodiments of the disclosure and does not pose a limitation on the scope of the disclosure unless otherwise claimed. No language in the specification should be construed as indicating any non-claimed element as essential to the practice of the disclosure.

[0178] All references, including publications, patent applications, and patents, cited herein are hereby incorporated by reference to the same extent as if each reference were individually and specifically indicated to be incorporated by reference and were set forth in its entirety herein.

[0179] In the description and claims, the terms "coupled" and "connected," along with derivatives thereof, can be used. It should be understood that these terms are not intended as synonyms for each other. Rather, in particular embodiments, "connected" or "coupled" can be used to indicate that two or more elements are in direct or indirect physical or electrical contact with each other. "Coupled" can also mean that two or more elements are not in direct contact with each other, but yet still co-operate or interact with each other.

[0180] Unless specifically stated otherwise, it can be appreciated that throughout the specification terms such as "processing," "computing," "calculating," "determining," or the like, refer to the action and / or processes of a computer or computing system, or similar electronic

[0181] In a similar manner, the term "processor" can refer to any device or portion of a device that processes electronic data from registers and / or memory to transform that electronic data into other electronic data that can be stored in registers and / or memory. As a non-limiting example, a "processor" can be a CPU or GPU. A "computing platform" can include one or more processors. As used herein, a "software" process can include, for example, software and / or hardware entities such as tasks, threads, and intelligent agents that perform work over time. Likewise, each process can refer to multiple processes to sequentially or concurrently execute instructions, either continuously or intermittently. The terms "system" and "method" can be used interchangeably herein so long as a system can embody one or more methods and a method can be considered a system.

[0182] In this document, obtaining, accessing, receiving, or inputting analog or digital data into a subsystem, computer system, or computer-implemented machine can be referenced. Analog and digital data can be obtained, accessed, received, or inputted in a variety of ways, such as by receiving data as a parameter of a function call or a call to an application programming interface. In some implementations, the process of obtaining, accessing, receiving, or inputting analog or digital data can be accomplished by transmitting data via a serial or parallel interface. In another implementation, the process of obtaining, accessing, receiving, or inputting analog or digital data can be accomplished by transmitting data from a providing entity to an accessing entity via a computer network. Providing, outputting, transmitting, sending, or presenting analog or digital data can also be referenced. In various examples, the process of providing, outputting, transmitting, sending, or presenting analog or digital data can be accomplished by transmitting data as an input or output parameter of a function call, a parameter of an application programming interface, or an interprocess communication mechanism.

[0183] Although the above discussion discloses example implementations of the described technology, other architectures can be utilized and are intended to fall within the scope of the present disclosure. Moreover, although a specific division of responsibilities has been defined above for purposes of discussion, various functions and responsibilities can be distributed and divided in different ways depending on circumstances.

[0184] Further, although the subject matter has been described in language specific to structural features and / or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described. Rather, the specific features and acts are disclosed as example forms of implementing the claims.

Claims

1. A computer-implemented method comprising: encoding one or more features of a first set of labeled synthetic data and one or more features of a second set of unlabeled real data to a latent space; providing the latent space as input to train a scene graph prediction network; aligning the one or more features in the latent space and one or more features in an output space of the scene graph prediction network; aligning one or more labels of the synthetic data with the real data; and training the scene graph prediction network using the one or more aligned labels.

2. The computer-implemented method of claim 1, further comprising: aligning the one or more features in the latent space and the one or more features in the output space using at least one of: one or more gradient reversal layers (GRLs) or a domain discriminator.

3. The computer-implemented method of claim 2, wherein aligning the one or more features reduces at least one of one or more disparities in appearance or content between the set of labeled synthetic data and unlabeled real data.

4. The computer-implemented method of claim 1, further comprising: aligning the one or more labels using pseudo-statistics based self-learning.

5. The computer-implemented method of claim 1, further comprising: receiving an image; and generating a scene graph for the image using the trained scene graph prediction network.

6. The computer-implemented method of claim 1, wherein the training includes applying network fusion criteria.

7. The computer-implemented method of claim 1, further comprising: generating a scene graph using the trained scene graph prediction network; and generating a synthetic image from the generated scene graph.

8. The computer-implemented method of claim 1, wherein the one or more features are encoded to the latent space using one or more convolutional neural networks (CNNs), the latent space including one or more features of both the first set of labeled synthetic data and the second set of unlabeled real data.

9. The computer-implemented method of claim 1, wherein aligning the one or more features reduces a disparity in appearance between the synthetic data and the real data, and wherein aligning the one or more labels reduces a disparity in content between the synthetic data and the real data.

10. A system comprising: at least one processor; and memory including instructions that, when executed by the at least one processor, cause the system to: encode one or more features of a first set of labeled synthetic data and a second set of unlabeled real data to a latent space; provide the latent space as input to train a scene graph prediction network; align the one or more features in the latent space and one or more features of an output space of the scene graph prediction network; align one or more labels between the synthetic data and the real data; and train the scene graph prediction network using the one or more aligned labels. ​ ​ ​ ​ ​ 11. The system of claim 10, wherein the instructions, when executed, further cause the system to: align the one or more features in the latent space with the one or more features in the output space using at least one of: one or more gradient reversal layers (GRLs) or a domain discriminator.

12. The system of claim 11, wherein aligning the one or more features reduces one or more differences in appearance and content between the set of labeled synthetic data and unlabeled real data.

13. The system of claim 10, wherein the instructions, when executed, further cause the system to: align one or more labels using self-learning based on pseudo statistics.

14. The system of claim 10, wherein the instructions, when executed, further cause the system to: receive an unlabeled image; and generate a scene graph for the image using a trained scene graph prediction network.

15. The system of claim 10, wherein the system comprises at least one of: a system for performing graphics rendering operations; a system for performing simulation operations; a system for performing simulation operations to test or validate autonomous machine applications; a system for performing deep learning operations; a system implemented using edge devices; a system that consolidates one or more virtual machines (VMs); a system implemented at least partially in a data center; or a system implemented at least partially using cloud computing resources.

16. A non-transitory computer-readable storage medium comprising instructions that, when executed, cause one or more processors to: encode one or more features of a first set of labeled synthetic data and one or more features of a second set of unlabeled real data into a latent space; provide the latent space as input to train a scene graph prediction network; align the one or more features in the latent space with the one or more features in an output space of the scene graph prediction network; align one or more labels of the synthetic data with the real data; and train the scene graph prediction network using the one or more labels.

17. The non-transitory computer-readable storage medium of claim 16, wherein the instructions, when executed, further cause one or more processors to: align the one or more features in the latent space with the one or more features in the output space using at least one of: a gradient reversal layer (GRL) or a domain discriminator.

18. The non-transitory computer-readable storage medium of claim 17, wherein aligning the one or more features reduces one or more differences in appearance and content between the set of labeled synthetic data and unlabeled real data.

19. The non-transitory computer-readable storage medium of claim 16, wherein the instructions, when executed, further cause one or more processors to: align the one or more labels using self-learning based on pseudo statistics. ​ 20. The non-transitory computer-readable storage medium of claim 16, wherein the instructions, when executed, further cause one or more processors to: receive an unlabeled image; and generate a scene graph for the image using the trained scene graph prediction network.

Images