221 results about "Object Class" patented technology

The present disclosure relates to systems, methods, and non-transitory computer readable media that generate inpainted digital images utilizing class-specific cascaded modulation inpainting neural network. For example, the disclosed systems utilize a class-specific cascaded modulation inpainting neural network that includes cascaded modulation decoder layers to generate replacement pixels portraying a particular target object class. To illustrate, in response to user selection of a replacement region and target object class, the disclosed systems utilize a class-specific cascaded modulation inpainting neural network corresponding to the target object class to generate an inpainted digital image that portrays an instance of the target object class within the replacement region. Moreover, in one or more embodiments the disclosed systems train class-specific cascaded modulation inpainting neural networks corresponding to a variety of target object classes, such as a sky object class, a water object class, a ground object class, or a human object class.

A text-to-image model training method, apparatus, and computer-readable storage medium for enhancing text-to-image generation through object-aware training. The method trains a text-to-image model using cyclic iterative training with sample image and text pairs. Training involves selecting image-text sample pairs containing multiple objects, obtaining corresponding mask images and object class names that distinguish location regions of the objects, and inputting both the sample image with description text and the mask images with object class names into the model. The method obtains image predicted noise and object predicted noises, constructs a loss function based on these predictions, and performs parameter adjustment accordingly. This approach enables improved object-level understanding in text-to-image generation models.

Systems and methods are provided for implementing training of a pre-trained object detection model for detecting new object classes. In examples, to train an object detection model, which has been pre-trained with a first set of object classes, with a new object class, a computing system applies to each of a plurality of first images that each depicts an object corresponding to an object class among the first set of object classes, a set of data augmentations combining each first image with at least one second image among a plurality of second images that each depicts a second object corresponding to the new object class, to generate a plurality of augmented images. The computing system trains the object detection model using the plurality of augmented images. In examples, original weights corresponding to the first set of object classes are retained, while random weights are used for the new object class.

The present application provides a method for device, and computer storage medium for generating an image caption. The method comprises: obtaining an image to be processed and auxiliary caption information, wherein the image to be processed includes a main object, and the auxiliary caption information includes at least one of the following: name information corresponding to the main object, object category corresponding to the main object, an object attribute corresponding to the main object, and an image tag corresponding to the image to be processed; determining an image feature corresponding to the image to be processed, and an auxiliary feature corresponding to the auxiliary caption information; generating the caption based on the image feature and the auxiliary feature to obtain a target caption corresponding to the image to be processed, wherein the target caption includes the name information of the main object.

A computer-implemented method for classifying points in a point cloud obtained by an active detection and ranging system is provided. The method comprises: computing a latent representation of the point cloud, determining from the latent representation, for each point in the point cloud, a probability that the point represents an object belonging to one or more object classes, and determining from the latent representation, for each point in the point cloud, a probability the point represents a ghost object. An object classification network is used for determining the probability that the point represents an object, and a ghost classification network is used for determining the probability the point represents a ghost object.

The present disclosure relates to systems, methods, and non-transitory computer-readable media that generate and implement behavioral policy graphs for digital image editing. For instance, in some embodiments, the disclosed systems generate, for a client device, a behavioral policy graph that assigns behaviors to object classes based on object relationships. The disclosed systems receive, from the client device, a digital image portraying a plurality of objects. Further, the disclosed systems determine behaviors of the plurality of objects utilizing the behavioral policy graph by determining, for each object of the plurality of objects, a behavior based on a relationship of the object with an additional object of the plurality of objects in accordance with the behavioral policy graph. The disclosed systems modify the digital image by modifying one or more objects based on the behaviors of the plurality of objects.

Approaches presented herein provide for the unconditional generation of novel three dimensional (3D) object shape representations, such as point clouds or meshes. In at least one embodiment, a first denoising diffusion model (DDM) can be trained to synthesize a 1D shape latent from Gaussian noise, and a second DDM can be trained to generate a set of latent points conditioned on this 1D shape latent. The shape latent and set of latent points can be provided to a decoder to generate a 3D point cloud representative of a random object from among the object classes on which the models were trained. A surface reconstruction process may be used to generate a surface mesh from this generated point cloud. Such an approach can scale to complex and / or multimodal distributions, and can be highly flexible as it can be adapted to various tasks such as multimodal voxel- or text-guided synthesis.

A computer-implemented method for building an object detection module uses mesh representations of objects belonging to specified object classes of interest to render images by a physics-based simulator. Each rendered image captures a simulated environment containing objects belonging to multiple object classes of interest placed in a bin or on a table. The rendered images are generated by randomizing a set of parameters by the simulator to render a range of simulated environments. The randomized parameters include environmental and sensor-based parameters. A label is generated for each rendered image, which includes a two-dimensional representation indicative of location and object classes of objects in that rendered image frame. Each rendered image and the respective label constitute a data sample of a synthetic training dataset. A deep learning model is trained using the synthetic training dataset to output object classes from an input image of a real-world physical environment.

The present application discloses an object category recognition model training method and apparatus, and an object category recognition method and apparatus, which can be applied to various scenarios such as cloud technology, artificial intelligence, intelligent transportation, and the Internet of vehicles. The method comprises: acquiring a newly added sample image of a sample object; on the basis of an initial object category recognition model, constructing a teacher model and a student model; inputting the newly added sample image into the teacher model for performing first object category recognition, and obtaining a first sample category; performing fusion on the first sample category and a newly added object category, and obtaining a fused category; inputting the newly added sample image into the student model for performing second object category recognition, and obtaining a second sample category; and, on the basis of a difference between the second sample category and the fused category, training the student model, so as to obtain a final object category recognition model. According to the model training method of the present application, the training duration is greatly shortened, and the model updating efficiency is improved.

Approaches presented herein provide for the unconditional generation of novel three dimensional (3D) object shape representations, such as point clouds or meshes. In at least one embodiment, a first denoising diffusion model (DDM) can be trained to synthesize a 1D shape latent from Gaussian noise, and a second DDM can be trained to generate a set of latent points conditioned on this 1D shape latent. The shape latent and set of latent points can be provided to a decoder to generate a 3D point cloud representative of a random object from among the object classes on which the models were trained. A surface reconstruction process may be used to generate a surface mesh from this generated point cloud. Such an approach can scale to complex and / or multimodal distributions, and can be highly flexible as it can be adapted to various tasks such as multimodal voxel- or text-guided synthesis.

The invention provides a knowledge base retrieval method and device, electronic equipment and a storage medium, and belongs to the technical field of data processing.The method comprises the steps that a target retrieval image is obtained and input into a target retrieval model, and the target retrieval model comprises a target visual large model and a target cross-modal mapping module; performing visual feature extraction on a target retrieval image through a target visual large model to obtain target retrieval visual features, and mapping the target retrieval visual features into text features through a target cross-modal mapping module to obtain target retrieval text features; according to the target retrieval visual features and the target retrieval text features, a pre-constructed multi-modal knowledge base is retrieved, the initial object category of the target object and the retrieval score of the initial object category are obtained, the target object category of the target object is determined according to the initial object category and the retrieval score, and the accuracy of knowledge base retrieval is improved.

A condition definition apparatus includes: accepting first inspection object data showing an image of a first inspection object; summary text data explaining a summary of condition of the inspection object in text, and detailed text data explaining details of condition of the inspection object in text, generating an image feature vector showing features of the image shown by the first inspection object data from the accepted first inspection object data, generating a summary feature vector showing summary of features of the condition of the inspection object shown by the summary text data from the accepted summary text data, generating an integrated feature vector by integrating the generated summary feature vector and generating the detailed feature vector and inspection object class data used to define the condition of the inspection object from the integrated feature vector.

The invention provides a text classification method and device, equipment and a storage medium, which can be applied to the field of maps and the like, and the method at least comprises the following steps: obtaining respective object information of each to-be-classified object, and extracting respective object text features of each obtained object information; on the basis of the obtained text features of the objects and sample features of pre-stored object samples, target samples matched with the object information are selected from the object samples; on the basis of the obtained target examples, the object information and the object categories, classification task texts corresponding to the to-be-classified objects are generated respectively; and adopting a pre-stored first large language model, based on the obtained classification task texts, reasoning the to-be-classified objects and corresponding target categories in the object categories, and obtaining a text classification result. The demanded quantity of the labeled training data is reduced, the problem that the data volume of the collected training data is unbalanced is avoided, and the classification accuracy during text classification is more stable.

The invention discloses a lightweight malicious software classification method based on multi-feature fusion, and the method comprises the following steps: extracting an operation code sequence and a static API call sequence from an ASM file of malicious software as original input, and carrying out the duplicate removal and length interception preprocessing of the sequences; performing multi-scale behavior feature extraction on the operation code sequence through an improved Res2Net module, analyzing the static API call sequence into an action-object-category triple semantic chain, and mapping the triple semantic chain into an embedded vector; splicing the obtained operation code feature vector and the API semantic feature vector into a fusion feature; after position codes are added to the fusion features, a multi-head self-attention mechanism is input, and key behavior feature representation is enhanced through attention weighting; the attention output is input into a full connection layer after pooling dimensionality reduction, and malicious software family classification is completed; according to the invention, the classification accuracy and generalization ability are effectively improved.

The invention discloses an object recognition method and device, equipment and a storage medium, and relates to the technical field of machine learning, and the method comprises the steps: carrying out the feature similarity analysis of a target object feature of a to-be-recognized object and a class index feature group corresponding to a target object class, and obtaining target feature similarity data; the category index feature group is a clustering center feature corresponding to the positive sample object feature; the positive sample object is a sample object which is screened based on a feature density clustering result of a plurality of original sample objects and belongs to a target object category; based on the target feature similarity data and a preset feature similarity threshold, determining object category indication information of the to-be-recognized object; the preset feature similarity threshold value is obtained by performing threshold value adaptive adjustment based on a preset identification index threshold value and the sample feature similarity data; the sample feature similar data represents the feature similarity condition between the positive sample object feature of the positive sample object and the category index feature group. By using the scheme of the invention, accurate identification of a new category can be rapidly realized.

A device and a computer implemented method for explainable clustering of a scene. The method includes determining a first relation that relates a first object class to a second object class, wherein determining the first relation includes determining, depending on the first object class and the second object class, a pair of entities in a first knowledge graph, in particular a commonsense knowledge graph, that represents information about a domain, wherein the pair of entities is related with the first relation in the first knowledge graph, determining a cluster in that the scene belongs depending on the scene and depending on other scenes, determining a second relation that relates the scene with the cluster depending on at least one feature of digital image data representing the scene, determining a rule that maps the first relation to the second relation.

In order to at least partially automatically control the braking of a vehicle (1), object data containing an object class (9) of an object (6) in an environment of the vehicle (1) is generated or received by means of a data processing system (3) of the vehicle (1). By means of the data processing system (3), environment data (10) which contains statistical traffic accident data relating to the environment and / or relating to environment conditions of the environment is generated or received. By means of the data processing system (3), control data relating to a braking intensity for a desired braking maneuver of the vehicle (1) is determined according to the object class (9) and the environment data (10), and at least one control signal for at least one brake actuator of the vehicle (1) is generated according to the control data.

In a method for training a neural network for detecting an object, geometric dimensions of a test object from an object class are captured, and during a time period, recordings of the test object are generated by a plurality of cameras. From the captured geometric dimensions and the generated recordings, occupancy maps are generated. By a radar device, a radar signal is transmitted, and a radar signal reflected by the test object is received. The transmitted radar signal and the received radar signal are mixed into a complex baseband to form a mixed signal. A complex four-dimensional mixed spectrum of the mixed signal is calculated. From the complex four-dimensional mixed spectrum, a first complex two-dimensional partial spectrum and a second complex two-dimensional partial spectrum are calculated. The occupancy maps and the partial spectra are fusioned to form training data. The training data are fed to the neural network.

Methods and systems are provided for inferring thickness and volume of one or more object classes of interest in two-dimensional (2D) medical images, using deep neural networks. In an exemplary embodiment, a thickness of an object class of interest may be inferred by acquiring a 2D medical image, extracting features from the 2D medical image, mapping the features to a segmentation mask for an object class of interest using a first convolutional neural network (CNN), mapping the features to a thickness mask for the object class of interest using a second CNN, wherein the thickness mask indicates a thickness of the object class of interest at each pixel of a plurality of pixels of the 2D medical image; and determining a volume of the object class of interest based on the thickness mask and the segmentation mask.

An autonomous driving system includes an object detection system. A neural network image encoder generates an image embedding associated with an image including an object. A neural network text encoder generates a concept embedding associated with each of a plurality of concepts. Each of the plurality of concepts is associated with one of at least two object classes. A confidence score module generates a confidence score for each of the plurality of concepts based on the image embedding and the concept embedding associated with the concept. An object class prediction module generates a predicted object class for the object based on an association between a concept set of the plurality of concepts having at least two of the highest values of the generated confidence scores and one of the at least two object classes associated with a majority of the concept set.

In a method for determining and classifying at least one object in a sensing area of a sensor, a two- or three-dimensional image of the sensing area is captured by means of the sensor. Sensor data corresponding to the image is generated. The object in the image is determined by means of the sensor data and by means of first template data of at least one first template object using a neural network. The neural network has been trained using the first template data. The first template data corresponds to an image of the first template object. At least one first object class is associated with the first template object. The object is classified by the neural network determining whether the object belongs to the first object class or not.

A method and system is disclosed for tuning a machine learning classifier. An object class requirement may be provided and include rank thresholds. The object class requirements may also include a range goal that defines a minimum distance from the object the machine learning algorithm should not provide false positive results. A base classifier may be trained using a weighted loss function that includes one or more weight values that are computed using the one or more object class requirements. An output of the weighted loss function may be evaluated using an objective function which may be established using the one or more object class requirements. The one or more weights may also be re-tuned using the weighted loss function if the output of the weighted loss function does not converge within a predetermined loss threshold.

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for object detection using neural networks. In one aspect, one of the methods includes obtaining an input image; processing the input image using an sequence transduction neural network to generate an output sequence that comprises respective token at each of a plurality of time steps, wherein each token is selected from a vocabulary of tokens that comprises (i) a first set of tokens that each represent a respective discrete number from a set of discretized numbers and (ii) a second set of tokens that each represent a respective object category from a set of object categories; and generating, from the tokens in the output sequence, an object detection output for the input image.

The embodiment of the invention discloses a data processing method and device, electronic equipment and a computer readable storage medium. According to the embodiment of the invention, at least one training data pair of a target detection model is obtained, the training data pair comprises a text sample and an image sample, then a text object is extracted from the text sample, at least one object anchoring area where the text object is located is extracted from the image sample, and then the target detection model is obtained based on the object anchoring area. Clustering the text objects to obtain at least one object category, and calculating category bias information of each object category; when a to-be-detected image is obtained, performing target detection on the to-be-detected image by adopting the target detection model to obtain an initial detection result, and then adjusting the initial detection result by adopting the category bias information; according to the scheme, the target detection accuracy can be improved; the embodiment of the invention can be applied to various scenes such as artificial intelligence, computer vision, natural language processing and the like.

A method and system perform incremental learning object detection in images and / or videos without catastrophic forgetting of previously learned object classes. A two-stage neural network object detector is trained to locate and identify objects belonging to additional object classes by iteratively updating the two-stage neural network object detector until an overall detection accuracy criterion is met. The update is performed to balance minimizing the loss of the initial ability to locate and identify objects belonging to previously learned object classes and maximizing the ability to additionally locate and identify objects belonging to additional object classes. Evaluating whether the overall detection accuracy criterion is met compares the output of an initial version of the two-stage neural network object detector with current region proposals output by a current version of the two-stage neural network object detector to determine a region proposal distillation loss and a previously learned object recognition distillation loss.

Included is a method for estimating distance, including: emitting, with at least one light emitter disposed on a robot, a light structure onto objects; capturing, with at least one image sensor disposed on the robot, images of the light structure emitted onto the objects; identifying, with a processor, the light structure within the images; determining, with the processor, positions of elements of the light structure within the images; determining, with the processor, a characteristic relating to the objects based on positions of elements of the light structure within the images; determining, with the processor, an object class of at least one object within the images based on a comparison between features of the object extracted from the images and an object dictionary comprising various object classes and their associated features; and instructing, with the processor, the robot to execute at least one action based on the object class identified.

Examples of the present disclosure describe systems and methods for sensory and response modeling in OWT systems. In examples, a payload is received by a sensory machine learning (ML) model implemented within an OWT system. The sensory ML model outputs an indication associated with data within the payload, such as whether the data belongs to one or more object classes or is indicative of anomalous activity. The output of the sensory ML model is provided to a response ML model implemented within the OWT system. The response ML model outputs a determination associated with the payload, such as whether the payload is permitted to egress across a data boundary of the OWT system or the manner in which data in the payload can be used in the one or more computing environments. The payload is then processed in accordance with the determination.

A computer system and method for training a machine learning system to perform a classification task by dividing input data into one of a plurality of classes. The system is configured to receive training data for each class from which a representation of each class can be derived, wherein each class is described by a plurality of representations; process the training data to form, for at least one class, a first proxy for a relatively global portion of training data items and a plurality of proxies for different relatively local portions of training data items, each proxy corresponding to a representation of data belonging to the class. For each training data item, the system is configured to: evaluate a match between the training data item and the proxies; estimate the class of the training data item from the match rating; adjust the proxies by updating a weighting matrix to reduce the distance between the training data item and the proxies of the estimated class. Defining a plurality of proxies in this way can result in richer and more stable representations of object classes.

Methods for automated process monitoring, wherein - image data (4) depicting a scene are generated by means of a camera system (2); - by means of a computing unit (3) based on the image data (4) it is determined that objects (6a, 6b, 7a, 7b, 7c, 8, 9, 10, 11, 14) are located within a specified area (B1, B2); - by means of the computing unit (3) for each of the objects (6a, 6b, 7a, 7b, 7c, 8, 9, 10, 11, 14) based on the image data (4) one of at least two predefined object classes is determined; - using the computing unit (3) depending on the specific object classes, it is checked whether a predefined rule assigned to the area (B1, B2) is fulfilled; - by means of the computing unit (3) an output signal is generated depending on a result of the check; - using the computing unit (3) based on the image data (4) for each of the objects (6a, 6b, 7a, 7b, 7c, 8, 9, 10, 11, 14) one of at least two predefined sub-areas (B1.1, B1.1.1, B1.2, B2.1, B2.1.1, B2.2, B2.2.1) of the area (B1, B2) is determined within which the respective object (6a, 6b, 7a, 7b, 7c, 8, 9, 10, 11, 14) is located; and - the verification of whether the rule is fulfilled is carried out depending on the specific sub-areas (B1.1, B1.1.1, B1.2, B2.1, B2.1.1, B2.2, B2.2.1).

The present invention relates to a computer-implemented method of training an object detector (OD), an object detector (OD), a computer program and a computer-readable (storage) medium. In order to carry out the method, a set of object classes (C) obfuscated by a neural network (NN), an object feature map (FM) originating from the neural network (NN), and object class tags (LO) comprising object classes (C) assigned to each object (O) in the object feature map (FM) need to be acquired. Subsequently, a training head (TH) configured to determine the authenticity of an object class (C) assigned to the object (O) in the object feature map (FM) from the object feature map (FM) and the object class tag (LO) is added to the neural network (NN). A modified object category tag (LO ') is determined from the set of object categories (C) and the object category tag (LO), wherein at least one object category assignment (C) is modified. The modified object category tags (LO ') are processed together with the object feature map (FM) by a neural network, generating an output, which is compared with an authenticity tag (LT) using a first objective function (OF1), the authenticity tag comprising an indication of the authenticity of each object category (C) in the modified object category tags (LO'). And updating the network parameters according to the comparison result.