8 results about "Mutual learning" patented technology

The steps of the EEG emotion recognition system based on contrast learning and implicit emotion regulation mechanism are as follows: first, the preprocessed EEG feature matrix is divided into left and right brain two-dimensional EEG feature matrices according to the left and right electrode distribution respectively; the obtained and a randomly initialized adjacency matrix are input into a dynamic connection EEG representation extraction module to obtain left and right brain shallow emotion representations and left and right brain deep emotion representations respectively, and an emotion classification loss is calculated. Then, the obtained and are input into an automatic reverse regulation module to calculate a contrast loss; the obtained and are input into a brain lateralization mutual learning module to calculate a KL loss. The total loss obtained from and is used to constrain the system, and EEG emotion recognition network parameters γ are obtained. Finally, the EEG data to be tested is input into the EEG emotion recognition network, and the final emotion recognition result is obtained using γ. The present application further improves the EEG emotion recognition accuracy from the perspective of biological mechanism.

The application provides a pyramid knowledge distillation framework model compression limit analysis method, comprising the following steps: constructing N groups of online deep mutual learning models in a pyramid structure; performing online deep mutual learning on each group of online deep mutual learning models, and recording the parameter quantity and model performance of two models in each group of online deep mutual learning models; wherein, starting from the second group of online deep mutual learning models from bottom to top, while performing online deep mutual learning, the previous group of online deep mutual learning models is accepted for offline knowledge distillation; the potential representation of all models from the first group to the N-1th group is extracted and sent to an adapter to generate teacher importance weight soft labels; the Nth group of online deep mutual learning models is subjected to online deep mutual learning, and the parameter quantity and model performance of the Nth group of models are recorded; and the balance point of the model compression ratio and accuracy is analyzed according to the parameter quantity and model performance of two models in each group of online deep mutual learning models and the parameter quantity and model performance of the Nth group of models.

This invention provides a hallucination detection method for a large language model, comprising: extracting a set of first intrinsic features from the original response and constructing a first intrinsic uncertainty representation; inputting the input to a response hallucination detector and outputting a first hallucination probability distribution; guiding the large language model to generate a self-judgment result from the original response; extracting second intrinsic features from the self-judgment generation process and constructing a second intrinsic uncertainty representation; inputting the input to a judgment hallucination detector and outputting a second hallucination probability distribution; constructing a logical relationship between the hallucination probability distributions based on the symbolic semantics of the self-judgment result and quantifying it as a logical loss; constructing a total loss function by combining the classification losses of the response hallucination detector and the judgment hallucination detector respectively, and performing joint optimization training. This invention also provides a hallucination detection device, storage medium, and electronic device for a large language model. Therefore, by constructing a dual-view detection framework and a mutual learning mechanism of logical constraints, this invention can achieve more accurate and robust hallucination detection.

This invention discloses a 2D and 3D collaborative learning method for MRI lumbar spine and intervertebral disc segmentation, belonging to the field of intelligent medical image analysis technology. The method performs anisotropic resampling, intensity normalization, and region clipping on lumbar spine MRI data, and then inputs these data into 2D and 3D segmentation branches respectively. Cross-dimensional information interaction is achieved through bidirectional mapping from slice to volume and from volume to slice. A dynamic gating mutual learning mechanism is constructed based on confidence, consistency, and axial continuity to suppress low-confidence pseudo-supervision noise. Furthermore, anatomical topological priors are introduced into the segmentation results to perform instance separation, segment labeling, and topological validity correction on the vertebral bodies and intervertebral discs. Experiments demonstrate that this method can significantly improve the segmentation accuracy of complex anatomical structures and effectively solve the boundary ambiguity problem in anisotropic data. This invention forms a 2D and 3D collaborative learning segmentation approach that is both innovative and interpretable.

The application belongs to the technical field of deep learning, and relates to an audio and video parsing method based on noise label learning, which comprises the following steps: preprocessing original audio and video to obtain a segment-level input sequence; constructing a mutual learning noise-resistant double-flow network; training the mutual learning noise-resistant double-flow network according to a training set; comparing the validation set indicators of two sub-networks in the trained mutual learning noise-resistant double-flow network, and taking the sub-network with the larger validation set indicator as an audio and video parsing model; and parsing through the audio and video parsing model according to a test set to obtain a video prediction result. The mutual learning noise-resistant double-flow network is composed of two sub-networks with the same structure but different initializations, a cross filtering mechanism is executed according to the clean masks generated by the two sub-networks during the training of the mutual learning noise-resistant double-flow network, and the dynamic confidence ratio is gradually reduced through a cosine strategy, so that the problems of high pseudo-label noise rate and easy overfitting noise in the existing audio and video parsing task are solved.

This invention discloses a deep mutual learning-based blood vessel image segmentation method, belonging to the field of medical image segmentation technology. The method includes: a dual-branch mutual learning segmentation model comprising a parallel Attention_Unet_EfficientV2 model and a Base_Unet_EfficientNet model; the deep mutual learning-based dual-branch mutual learning segmentation model performs feature alignment and error correction between the Attention_Unet_EfficientV2 model and the Base_Unet_EfficientNet model through collaborative training; the Attention_Unet_EfficientV2 model is constructed by incorporating an attention mechanism on the basis of the traditional U-Net model; the Base_Unet_EfficientNet model is an improvement on the traditional U-Net model and EfficientNet-B5; the deep mutual learning-based dual-branch mutual learning segmentation model is used to segment the blood vessel image to be segmented, obtaining the blood vessel contours in the image. This method improves the segmentation accuracy of blood vessel images.

The application discloses a tongue disease risk prediction method and system based on prompt mutual learning. The method comprises the following steps: training a visual language teacher model based on a tongue data set through prompt learning and consistency loss, and saving a text feature vector generated by the visual language teacher model; initializing at least two student models for mutual learning, and introducing a KL divergence loss of the visual language teacher model in the mutual learning; and multiplying an image feature vector of a tongue image to be predicted extracted by any one student model with the text feature vector generated by the visual language teacher model to output a disease risk prediction result. The application effectively improves the accuracy and reliability of tongue disease risk prediction by fusing prompt learning and mutual learning distillation technology.

The present application belongs to the technical field of intelligent processing of satellite-borne remote sensing, and particularly relates to a teacher-supervised mutual learning knowledge distillation satellite-borne remote sensing scene classification method based on standardized contrast learning. The specific process is as follows: the output of the teacher-student classification network is standardized by using RobustScaler, and the SCL loss is calculated; the mutual learning loss of N lightweight student models is calculated; the projector is set to ensure feature alignment, and the bridge loss is calculated; the TSML total loss is set based on the above losses, and the hyperparameters of each loss are set; during training, each hyperparameter of the TSML total loss is fine-tuned and optimized based on the greedy genetic algorithm; and the satellite-borne remote sensing scene is classified by using the trained teacher-student classification network.