25 results about "Discriminative learning" patented technology

Examples of implementation and training of artificial intelligence models developed with self-supervised learning for medical image analysis are disclosed. The models as trained leverage training from discriminative learning, restorative learning, and adversarial learning in a unified manner to glean complementary visual information from unlabeled medical images for fine-grained semantic representation learning. The models as trained provide features suitable for generalizable representation of medical input images many organs, diseases, and modalities.

The invention provides a rotating machine transfer learning fault diagnosis method fusing semi-supervised contrast learning in the field. In the method, an intra-domain semi-supervised contrast learning (SSCL) algorithm is designed, the SSCL takes category information as supervision, discriminative learning of different categories in each domain is guided to eliminate fuzzy classification boundaries, convenience is provided for domain adaptation of cross-domain features carried out by adopting local maximum mean difference (LMMD), and then cross-working-condition diagnosis performance is improved. Meanwhile, domain confrontation is introduced in a mode of dynamically limiting related contrast learning loss and transfer learning loss gains, so that negative effects of target domain pseudo labels with poor quality on SSCL and feature transfer learning are reduced, and the stability of a diagnosis model is improved.

The invention discloses a weak supervision time sequence action positioning method based on mask pyramid enhancement, and belongs to the field of computer vision and action positioning. A system module for realizing the method comprises a feature extraction module, a mask pyramid classification module MPCM for motion integrity learning, and an attention module FDAM for feature discrimination learning. The feature extraction module is used for extracting features as input data of subsequent links; the MPCM constructs a refined and enhanced mask pyramid by using a time sequence adaptive enhancement mechanism (TAEM), and introduces a mask strategy to drive a model to mine a new feature region complementary with a target action during hierarchical processing. And performing perceptual fusion on classification results of different levels to generate an enhanced class activation sequence (CAS). The MPENet introduces mutual exclusion loss in FDAM to help the model focus on the key part of the target action, so that the model is guided to learn proper attention distribution, and the understanding and distinguishing ability of the model to features is improved. And finally, weighting the CAS and the attention to obtain a prediction result of the model.

The invention discloses an RGB-D salient target detection method based on decoupling contrast learning, and designs a saliency detection framework integrating expression enhancement, modal collaborative perception and structural discrimination learning by combining a structural heterogeneity problem in multi-modal modeling and utilizing the frequency domain structural advantage of a deep mode and the long-distance modeling capability of Transform. By introducing wavelet convolution and Transform joint modeling, a cross-modal interaction parallel fusion mechanism and a pixel-level structure perception contrast learning strategy, high-precision, multi-scale and boundary clear detection of a salient target area in a complex scene is realized. The method can effectively solve the problems of large information difference between modes of the RGB and the depth map, difficulty in structure alignment, fuzzy boundary prediction, weak feature expression ability and the like, significantly improves semantic consistency and structural integrity of the salient region, and has good cross-modal generalization ability and robustness.

The invention relates to an anti-noise semantic communication method based on supervised discriminant learning, and belongs to the technical field of communication. For the problems of semantic noise and channel noise in image transmission, a deep supervision network is introduced at a semantic task execution end, a supervision layer is additionally arranged behind a hidden layer, and high-quality middle layer features are extracted, so that semantic feature representation is learned; a composite loss function containing terminal loss and adjoint loss is designed, the adjoint loss is introduced into a prototype consistent loss function, and semantic noise influence is reduced through maximum separation constraint; according to the method, the priority of task execution and data recovery is adjusted according to the SNR value so as to adapt to different communication conditions; furthermore, semantic noise is generated by adopting a white box attack method, and the anti-noise performance is enhanced by utilizing a supervision layer output result on the premise of not changing the structure of the original deep learning model; a loss function is designed from the semantic feature level, the anti-noise performance is further enhanced, and data recovery and task execution are both considered.

The invention discloses a predicate similarity discrimination method integrating sensory and perceptual clues, and the method comprises the following specific steps: extracting appearance features, spatial features and entity tags of an input image in a target detection process, and combining target pairs to construct joint features; and then the sensory sensitive module is responsible for capturing intuitive visual clues in the image as sensory clues, aggregating entity-level visual information, and realizing coarse-grained screening and feature alignment of candidate entity pairs. And the perception enhancement module further models the deep relationship between the entity pairs on the basis of the output of the perception module, and realizes semantic differentiation and optimization of similar predicates through semantic mapping and discriminative learning so as to conclude perception clues. And finally, performing deep association on visual scenes with a unified relationship in the same region by the distinguished triple category in a cognitive discrimination layer in a perception enhancement module, and ensuring that predicates embedded between entity pairs can capture discovered fine-grained semantics on the premise of not influencing semantics.

A Discriminative, Restorative, and Adversarial (DiRA) learning framework for self-supervised medical image analysis is described. For instance, a pre-trained DiRA framework may be applied to diagnosis and detection of new medical images which form no part of the training data. The exemplary DiRA framework includes means for receiving training data having medical images therein and applying discriminative learning, restorative learning, and adversarial learning via the DiRA framework by cropping patches from the medical images; inputting the cropped patches to the discriminative and restorative learning branches to generate discriminative latent features and synthesized images from each; and applying adversarial learning by executing an adversarial discriminator to perform a min-max function for distinguishing the synthesized restorative image from real medical images. The pre-trained model of the DiRA framework is then provided as output for use in generating predictions of disease within medical images.

A stepwise incremental pre-training for integrating discriminative, restorative, and adversarial learning in an AI model. Exemplary systems include means for receiving a training dataset for training a unified AI model. The unified AI model applies each of discriminative, restorative, and adversarial learning operations through three transferable components: a discriminative encoder, a restorative decoder, and an adversarial encoder. Stepwise incremental pre-training operations train the unified AI model, including pre-training the discriminative encoder via discriminative learning and attaching the trained discriminative encoder with the restorative decoder to form a skip-connected encoder-decoder, pre-training the skip-connected encoder-decoder via joint discriminative and restorative learning, and associating the pre-trained skip-connected encoder-decoder with the adversarial encoder. Training of the AI model is finalized by performing full discriminative, restorative, and adversarial learning on the training dataset using the unified AI model.

This invention proposes an image multi-task recognition method and system based on advanced learning, constructing an end-to-end advanced multi-task deep learning network model for simultaneous facial attribute recognition. This advanced multi-task learning network consists of three main modules: a primary network, an advanced network, and an expert-level network. The primary network is constructed using a lifelong learning strategy, training on facial attribute samples to acquire basic facial attribute recognition capabilities. The advanced network is constructed using a multi-task learning strategy, training on specially collected facial multi-attribute samples for multi-task collaborative learning, enabling advanced recognition of multiple facial attributes simultaneously. The expert-level network is constructed using a soft-label learning strategy, training on specially collected, confidence-based facial multi-attribute training samples for discriminative learning, enabling expert-level identification capabilities that see beyond appearances.

The application relates to the technical field of natural languages, and discloses a question generation method, which comprises the following steps: acquiring a question data set, wherein the question data set comprises initial questions; generating questions by using a question generator to obtain candidate questions; generating answers by using an answer generator to obtain candidate answers and question answers; determining first similarity values and second similarity values according to all question answers corresponding to the same initial question and candidate answers corresponding to the same candidate question; and screening all candidate questions according to all first similarity values and all second similarity values to obtain target questions. Through discriminative learning on the adversarial questions, the learning result is fed back to the question generator for optimization, so that the question generation model has a higher accuracy in generating questions, and the accuracy of artificial customer service in replying to customer questions in the insurance field or the financial field is improved.

The application provides a lung cancer intelligent early screening system, comprising a data enhancement module, a discriminative feature extraction module and a fine-tuning module; the data enhancement module is used for generating weak enhancement samples and strong enhancement samples; the discriminative feature extraction module is used for extracting corresponding strong and weak enhancement sample feature vectors and classification vectors, and performing contrast learning and discriminative learning on the two respectively, narrowing the distance between samples related in semantics and excluding samples irrelevant in semantics; the fine-tuning module is used for optimizing the feature extractor, and the contrast learning of the original image of the labeled data and the corresponding weak enhancement sample is used to strengthen the accuracy of model classification.

The application provides a multi-modal emotion recognition method and system based on discriminative learning. The method comprises the following steps: step 1: collecting multi-modal information, including electroencephalogram signals, facial signals, speech signals and text signals; step 2: inputting the multi-modal information into a feature extraction network respectively to obtain electroencephalogram features, facial features, speech features and text features; step 3: using a canonical correlation analysis method to calculate the correlation between any two modal features; step 4: inputting the modal features into respective corresponding single-modal classifiers respectively to obtain single-modal prediction results; step 5: using the correlation between any two modal features and the single-modal prediction results to design a class loss function corresponding to each modal; step 6: obtaining a target function according to the class loss function corresponding to each modal to guide the training of an emotion recognition model; and step 7: obtaining an emotion recognition result by using the trained emotion recognition model according to multi-modal information of an object to be recognized.

The present application provides a method for spam comment detection based on an enhanced multi-relational graph neural network, which solves the technical problem that class imbalance and the "disguise" behavior of spam comments lead to poor detection model training, which in turn causes unsatisfactory detection results. The method comprises the following steps: obtaining the features of user comments and converting them into feature vector form; constructing a multi-relational graph of user comments with user comments as nodes, and dividing the graph into training and test sets; calculating the neighborhood homogeneity and label perception score of the training set to perform neighbor sampling; multiplying a random variable with the feature vector of an abnormal node in the training set to generate a new abnormal node; training a graph neural network model to perform discriminative learning on the nodes in the training set; using the trained graph neural network model to predict the nodes in the test set and output the prediction results. The present application is applied to the technical field of spam comment detection.

The invention provides an intelligent early screening system for lung cancer. The intelligent early screening system comprises a data enhancement module, a discriminative feature extraction module and a fine adjustment module, the data enhancement module is used for generating a weak enhancement sample and a strong enhancement sample; the discriminative feature extraction module is used for extracting corresponding strong and weak enhanced sample feature vectors and classification vectors, respectively performing comparative learning and discriminative learning on the strong and weak enhanced sample feature vectors and classification vectors, shortening the distance between semantically related samples and excluding semantically unrelated samples; and the fine tuning module is used for optimizing the feature extractor and enhancing the accuracy of model classification through comparative learning of an original image with label data and a corresponding weak enhancement sample.

The present invention discloses a robust image clustering method based on discriminative embedded projected fuzzy clustering, which belongs to the field of image recognition and classification and pattern recognition. The present invention adopts a robust image clustering method based on discriminative embedded projected fuzzy clustering, embeds the optimal subspace projection learning into the fuzzy clustering algorithm for optimization, and embeds discriminative learning while removing redundant features, effectively enhancing the inter-class discriminability of samples in the projected optimal subspace, thereby suppressing the degradation of clustering performance under noise pollution. In addition, inspired by the entropy metric, a regularization term based on the maximum entropy principle is designed, and a dynamic information entropy graph is constructed to update the sample membership level distribution to mine a more reliable natural category division of the data. The present invention adaptively identifies noise-contaminated image data while robustly clustering, and eliminates its influence. It has high robustness to noise-contaminated image data and effectively improves the clustering performance of noise-contaminated data.

The invention relates to the technical field of artificial intelligence, and provides a visual basic model training and visual feature extraction method, device and equipment, and the method comprises the steps: carrying out the training of a visual basic model based on the reconstruction prediction result of each image, the original image information of each image, and the classification prediction result and the classification false label of each image; and performing parameter iteration on the visual feature extraction model, and taking the initial visual basic model after parameter iteration as a visual basic model. According to the method, a classification task of discriminant learning and an image reconstruction task of generative learning are combined in a unified training framework, so that the visual basic model obtained by training retains rich local structures and detail representations learned by a generative model, and the accuracy of image reconstruction of the visual basic model is improved; and meanwhile, the high-level semantic understanding capability of the discriminant of the visual basic model is also improved, so that the method has excellent few-sample migration performance in tasks such as image classification and the like, and the deployment cost of the visual basic model is reduced.

The invention relates to an original monitoring video-oriented unsupervised pedestrian search method, which comprises the following steps of: detecting a pedestrian area in an original monitoring video by adopting a pre-trained neural network, and generating pedestrian features by adopting a feature encoder; in an original monitoring video, based on visual similarity and context similarity of pedestrians, searching for a first neighbor relation of pedestrian features and organizing a chain structure of the pedestrian features, determining spatial distribution of the pedestrian features, realizing unsupervised clustering according to segmented regions, and generating identity categories and pseudo labels of the pedestrians; in a forward propagation process, respectively calculating a contrast loss function of each branch from three branches of category distinguishing learning, single-camera contrast learning and multi-camera contrast learning, and respectively updating a clustering center memory bank and a multi-camera memory bank according to a momentum updating mechanism; and performing joint optimization on the contrast loss functions of the three branches, minimizing a total loss value through multiple iterations in a back propagation process, and optimizing parameters of the feature encoder.

The application relates to a scene-dependent video anomaly detection method and device. The method comprises: acquiring events and scene features. After feature preprocessing, random disturbance is applied to the scene features, and the event features are updated through time dimension self-attention. Modulation parameters are generated through a scene condition fusion block, the event features are modulated and fused, and residual connection is performed, a scene condition classification head is input, and discriminative learning is performed with a camera ID as a label. Meanwhile, the features are processed for comparative learning, positive and negative sample pairs are constructed in combination with a memory bank, and bidirectional symmetric comparative learning is performed. Two types of anomaly scores are calculated, adaptive fusion is performed based on distribution difference and local correlation, the maximum fusion score of each frame is taken as a frame-level score, and output is performed after median filtering and smoothing, so that accurate anomaly detection is realized. The method can improve the detection accuracy and practicability in a complex multi-scene environment.

The system receives a plurality of medical images and integrates Self-Supervised machine Learning (SSL) instructions for performing a discriminative learning operation, a restorative learning operation, and an adversarial learning operation into a model for processing the received plurality of medical images. The model is configured with each of a discriminative encoder, a restorative decoder, and an adversarial encoder. Each of the discriminative encoder and the restorative decoder are configured to be skip connected, forming an encoder-decoder. Step-wise incremental training to incrementally train each of the discriminative encoder, the restorative decoder, and the adversarial encoder is performed, in particular: pre-training the discriminative encoder via discriminative learning; attaching the pre-trained discriminative encoder to the restorative decoder to configure the encoder-decoder as a pre-trained encoder-decoder; and training the pre-trained encoder-decoder of the model using joint discriminative and restorative learning. The pre-trained encoder-decoder is associated with the adversarial encoder. The pre-trained encoder-decoder associated with the adversarial encoder is trained through discriminative, restorative, and adversarial learning to render a trained model for the processing of the received plurality of medical images. The plurality of medical images are processed through the model using the trained model.

The invention discloses a point supervision time sequence motion positioning method based on cooperation of motion discrimination and boundary sensitivity, and the method introduces a discrimination guide propagation mechanism, and definitely models discrimination fragment and fuzzy fragment semantic information. The method promotes knowledge propagation of the discrimination information to the fuzzy region and inhibits noise diffusion of the fuzzy region. Meanwhile, in combination with center contrast loss and feature consistency loss constraints, the discrimination of fragment features is enhanced, and a high-quality class activation score CAS is generated. In order to overcome the limitation of the CAS in the boundary modeling aspect, a boundary-sensitive time sequence modeling mechanism is designed to enhance the sensitivity of the boundary and the robustness of proposal expression. The action features are modeled in order, the structural expression of the action features is enhanced, and meanwhile potential association across time periods is captured in combination with disordered processing. On the basis, feature discriminative learning of the fragments to proposed boundary sensitivity enhancement is achieved, and then action positioning is improved.

In various examples, a technique for performing unified generative and discriminative learning includes converting, via execution of a machine learning model, a plurality of training data samples into a first plurality of latent representations. The technique also includes computing one or more losses based on the plurality of latent representations, wherein the loss(es) include a contrastive term that approximates an expected similarity between a latent representation of a training data sample and a second plurality of latent representations associated with a distribution of training data samples that includes the plurality of training data samples. The technique further includes updating one or more parameters of the machine learning model based on the one or more losses to generate a trained machine learning model.

The application discloses a weakly supervised temporal action localization method based on mask pyramid enhancement, and belongs to the fields of computer vision and action localization. System modules for realizing the method include a feature extraction module, a mask pyramid classification module MPCM for action integrity learning, and an attention module FDAM for feature discriminative learning. The feature extraction module is used for feature extraction and serves as input data of subsequent links; the MPCM constructs a refined enhanced mask pyramid by using a temporal adaptive enhancement mechanism TAEM, and introduces a mask strategy driven model in hierarchical processing to mine new feature regions complementary to a target action. Perception fusion is performed on classification results of different levels to generate an enhanced class activation sequence CAS. The MPENet introduces an exclusive loss in the FDAM, helps the model focus on key parts of the target action, thereby guiding the model to learn appropriate attention distribution, and improving the understanding and distinguishing ability of the model to features. Finally, the CAS is weighted with attention to obtain a prediction result of the model.

The application provides a layered robust discriminant learning method for hyperspectral remote sensing image anomaly detection. 1，1 The application effectively depicts the complex mixed noise introduced in the process of acquiring the hyperspectral remote sensing image in a real scene through the l norm and the Frobenius norm, and improves the anti-noise performance of the anomaly target detection model. In order to more accurately separate the deeply mixed background and anomaly target, obtain a robust and more powerful anomaly target detection model, and design a layered detection idea, the background and anomaly target components in the deeply mixed hyperspectral remote sensing image are gradually separated. The application can not only improve the distinguishability between the background and the anomaly target, but also has very strong noise suppression performance and detection robustness.

The system receives a plurality of medical images and integrates Self-Supervised machine Learning (SSL) instructions for performing a discriminative learning operation, a restorative learning operation, and an adversarial learning operation into a model for processing the received plurality of medical images. The model is configured with each of a discriminative encoder, a restorative decoder, and an adversarial encoder. Each of the discriminative encoder and the restorative decoder are configured to be skip connected, forming an encoder-decoder. Step-wise incremental training to incrementally train each of the discriminative encoder, the restorative decoder, and the adversarial encoder is performed, in particular: pre-training the discriminative encoder via discriminative learning; attaching the pre-trained discriminative encoder to the restorative decoder to configure the encoder-decoder as a pre-trained encoder-decoder; and training the pre-trained encoder-decoder of the model using joint discriminative and restorative learning. The pre-trained encoder-decoder is associated with the adversarial encoder. The pre-trained encoder-decoder associated with the adversarial encoder is trained through discriminative, restorative, and adversarial learning to render a trained model for the processing of the received plurality of medical images. The plurality of medical images are processed through the model using the trained model.

The application discloses a human action recognition method based on semantic perception fusion and adaptive discriminative learning, first action description semantic generation and coding; then video feature coding; finally, based on an adaptive mechanism and a contrast learning classifier, human action recognition is finally realized. The application solves the problem that the existing method based on an RGB video is often limited by background interference, motion blur and semantic inconsistency caused by modal heterogeneity, so that the recognition performance is difficult to meet the actual demand.