28results about How to "Reduce training overhead" patented technology

The invention relates to the technical field of emotion recognition, in particular to a multi-modal emotion recognition method based on deep learning, which comprises the following steps: constructing and training a multi-modal emotion recognition model, and inputting to-be-recognized data into the trained multi-modal emotion recognition model to output a recognition result; the multi-modal emotion recognition model comprises a self-attention encoder, a first cross attention encoder, a second cross attention encoder, a gating unit and an MLP classifier; according to the method, comprehensive learning and interaction are carried out by combining two modal data of voice and text, the emotion information is more comprehensive, the inherent defect of a single feature is made up, and the robustness of the model is higher and the accuracy is higher.

The invention relates to an RIS auxiliary safety wireless communication system based on a channel knowledge map and an environment perception beam selection method. The invention relates to an RIS auxiliary secure wireless communication environment sensing beam selection method based on a channel knowledge map, which reduces the training overhead on the premise of ensuring the RIS auxiliary communication secrecy rate: considering an RIS auxiliary secure communication system, and obtaining position information and beam pairs corresponding to users in a region by using an environment sensing technology; using an abstract function to link a beam pair in a given environment with a user position, and storing the beam pair and the user position in a CKM; the method comprises the following steps: for a downlink of an RIS assisted secure communication system, selecting a required optimal beam pair from a CKM under the condition that the positions of a BS and an RIS are fixed; an eavesdropper is considered in a communication environment, the selected beam pair is input into an alternating optimization AO algorithm, and the alternating optimization AO algorithm is adopted to carry out iterative optimization on the beam to obtain a final beam pair; the algorithm complexity and the number of iterations are analyzed, and the optimization rate is improved.

The invention discloses a facial acne grading method based on collaborative mapping multi-scale self-distillation, is applied to the field of computers, and aims to solve the problem of low accuracy of acne severity grading in the prior art. According to the method, the convolutional neural network model is trained by adopting a self-distillation architecture, and a novel distillation module optimization training process is provided in combination with the characteristics of acnes, so that the model can notice the uniqueness of acne images, and the severity grading precision of the model for facial acnes is improved. The method can automatically identify the acne features and grade the severity of the acne features, assists doctors in diagnosis and treatment, and accelerates the healing period of the acne.

The invention provides a method for determining the position of a temperature instrument of a solid-state reactor core and a method for training a temperature field prediction model. The method for determining the position of the temperature instrument comprises the following steps: determining a coordinate range in which the temperature instrument is allowed to be arranged according to a solid-state reactor core design scheme; in the coordinate range, p instrument coordinates are selected according to system sampling, and p is an integer larger than 1; calculating the importance degree of each instrument coordinate; and an instrument with high importance is selected as the arrangement position of the temperature instrument. The training method of the temperature field prediction model comprises the step that a training loss function comprises a regression loss item for a reactor core temperature field feature vector and a classification loss item for judging whether safety protection is triggered or not. Aiming at the problems of large temperature gradient and limited measurable position of a solid reactor core, the invention provides a method for evaluating the instrument coordinate importance degree through a deep learning neural network model, the instrument coordinates which are relatively important for temperature field prediction are screened out, and a corresponding temperature field prediction model is trained and formed.

The invention is suitable for the technical field of visual language models, and provides a distillation learning method of an open vocabulary target detection model, which comprises the following steps: based on a pre-trained candidate frame extraction model, obtaining a pseudo tag of a to-be-detected picture, cutting the to-be-detected picture according to a pseudo tag region, obtaining a pseudo tag region image, and obtaining an open vocabulary target detection model; the pseudo label is a candidate frame which is not overlapped with the label frame of the known class; based on a CLIP visual encoder, obtaining visual features of the pseudo label area image; on the basis of a CLIP text encoder, a text category label generation method is adopted, and text features of the pseudo label area image are obtained; and based on the visual features and the text features, performing knowledge distillation learning on the target detection model by adopting a bisymmetric distillation method, so that the target detection model learns the visual features of the CLIP visual encoder and the text features of the CLIP text encoder. And the generalization ability of the target detection model is improved.

The invention provides a cross-modal image-text retrieval method for the professional field, and the method comprises the steps: constructing an adapter according to an efficient parameter fine tuning technology, and training the adapter through a contrast learning training mode; according to the trained adapter, feature extraction and extension are carried out on image-text data, uploaded by a user, in different professional fields, and feature embedding vectors of expanded texts and images are utilized to construct an index structure based on a hierarchical graph; selecting a retrieval interface of a corresponding professional field according to query content of a user, and selecting a trained adapter of the corresponding field to perform feature extraction on the query content; and carrying out approximate nearest neighbor search on the extracted characteristics of the query content in the established index structure to obtain a most relevant query result, and returning the most relevant query result to the user. Based on the scheme provided by the invention, while accurate and efficient cross-modal retrieval service aiming at data in different professional fields is provided for the user, the training overhead is relatively low.

The invention discloses an edge terminal-oriented memory efficient model fine tuning method and system, which realize decoupling of a backbone network and a fine tuning process by constructing a parallel side network module, and avoid video memory occupation caused by backbone gradient return. A double-adapter module is arranged, modeling is carried out on same-layer features and cross-layer context information, and the feature expression precision is improved. And constructing a feature fusion module, and performing adaptive weighting on different path outputs through learnable gating. A backbone grouping module is integrated, dynamic grouping is carried out on a backbone network based on interlayer similarity, and redundant modules and calculation overhead are reduced. And through the fine tuning execution module, only back propagation and parameter updating are carried out on the side network, and a trunk freezing state is kept, so that the memory and training cost is reduced. According to the method, the memory efficiency and the training speed of the edge fine adjustment process are improved, high-precision model self-adaption is achieved with extremely low calculation burden, and application in the fields of cloud edge collaborative intelligence and the like of unmanned aerial vehicles, robots, vehicle-mounted terminals and the like is effectively supported.

The invention provides a large-model multi-language enhancement training method capable of continuously expanding experts, which comprises the following steps: step 1, extracting an original FFN module in a large-scale language model LLM to be trained and segmenting the original FFN module into a plurality of new fine-grained FFN modules as fine-grained experts; 2, setting a fine-grained router, and selecting a preset number of newly-added fine-grained experts according to a preset rule to obtain the output of the fine-grained router; 3, setting a parent router, controlling the weight between the output of the original FFN module and the output of the fine-grained router, and calculating the final output; 4, freezing an original FFN module, and independently training the fine-grained expert; step 5, freezing an original FFN module, adding the newly added fine-grained expert, and training the newly added fine-grained expert; step 6, freezing all other parameters, and only training a parent router; and 7, continuously repeating the step 4 to the step 6 to realize large-model multi-language enhancement training of sustainable extension experts.

A method, system, and apparatus for spectrum sharing in a massive multiple-input multiple-output (MIMO) network are disclosed. According to one aspect, there is provided a method in a secondary network node of a secondary network, the secondary network node being configured to communicate with a plurality of secondary users. The method includes performing channel estimation of a primary user of the primary network during a learning phase consistent with a training phase of the primary network. The method further includes determining a beamformer and a power allocation based at least in part on the channel estimate to maximize at least one of a weighted uplink data rate and a weighted downlink data rate under constraints with respect to a data rate of the primary network.

This invention belongs to the interdisciplinary field of artificial intelligence and cheminformatics, and provides a method, inference method, and apparatus for constructing a molecular multimodal large language model based on parameter space alignment. The method includes: constructing an instance set; constructing a molecular multimodal large language model, including an adaptive weight generator, a graph neural network encoder, and a large language model; training the molecular multimodal large language model using the instance set to update the parameters of the adaptive weight generator; the graph neural network encoder extracting node-level feature representations of the molecular structure diagram; the adaptive weight generator obtaining the low-rank weight update amount of the target weight component in the large language model based on the node-level feature representation, and injecting the low-rank weight update amount into the original weights of the target weight component; and the large language model after injecting the low-rank weight update amount generating an actual text response based on the input training text command. By integrating the molecular structure into the parameter layer of the large language model, deep perception of the molecular structure is achieved without changing the length of the input sequence.

The application discloses an improved ELM speech enhancement method and device for noise reduction, and belongs to the field of speech signal processing; specifically, first, a noisy speech signal and a clean speech signal are acquired to construct training samples; meanwhile, an improved ELM speech enhancement network is constructed, the training samples are used for training, and a network weight matrix is saved; then, a new noisy speech signal is acquired, input into the trained enhancement network to output a noise-reduced enhanced speech signal; the device comprises an audio acquisition module, an offline training module, an online running module and an audio playing module; the audio acquisition module acquires speech signal samples used for training and inputs the offline training module to train the improved ELM speech enhancement network and save the network weight matrix; a new noisy speech signal is acquired and input into the online running module to directly output a noise-reduced enhanced speech signal; and the audio playing module is used for playing. The application reduces training cost and has lower processing time delay.

The present invention discloses a method for acquiring channel states and target parameters of large-scale MIMO with integrated communication and perception. By parameterizing high-dimensional communication channels into a small number of physical parameters, an intrinsic correspondence with the perceived target parameters is established from the angle, delay and Doppler dimension. A shared pilot mode is proposed under the same time-frequency resources, so that channel estimation and target parameter estimation can be expressed in the same mathematical form as a structured tensor multiple decomposition model. Based on this model, the uniqueness conditions of the tensor decomposition problem and the maximum number of solvable targets are analyzed using the Vandermonde structure of the factor matrix. On this basis, a unified channel and target estimation algorithm is proposed to jointly estimate parameters including arrival angle / departure angle, delay, Doppler frequency shift, channel coefficient / reflection coefficient. The unified channel state and target parameter acquisition method proposed by the present invention has great advantages in reducing training overhead and improving perception resolution.

The invention relates to a multi-mode millimeter wave beam prediction method based on multi-task learning, and belongs to the technical field of millimeter wave communication. The method comprises the following steps: converting a beam prediction task into a deep learning optimization task based on a geometric channel model in a dynamic communication scene; preprocessing the image, the three-dimensional point cloud and the user motion information acquired by the base station to complete region-of-interest extraction, point cloud downsampling and space coordinate conversion; through a cross-modal gating fusion module, adaptively extracting and carrying out weighted fusion on the multi-modal features; and constructing a multi-task learning framework, cooperatively training a beam prediction main task and blocking prediction and reflection intensity prediction auxiliary tasks, correcting an optimal beam probability by using physical constraint information output by the auxiliary tasks, and selecting an optimal beam. According to the method, the beam training overhead and the communication delay are remarkably reduced, the environmental limitation of single-mode sensing is effectively overcome, and the prediction robustness of the system in a complex dynamic scene is enhanced while the beam prediction precision is improved.

The invention relates to the technical field of natural language processing, and discloses a method for improving context learning of a large language model based on parameter parsing, comprising: forming a verification prompt prefix from a task example sample and a verification context through a template of a specific task and inputting the prefix into a large language model to obtain a predicted verification context; dividing the large language model into group representations, observing the degree of performance degradation of the large language model on a specific task after being disturbed through an evaluation index, and establishing an association between the specific task and the group representation; improving the ability corresponding to the group representation through a representation improvement module; forming a test prompt prefix from a task example sample and a test context through a template of a specific task, and inputting the test prompt prefix into a large language model introduced with the representation improvement module to obtain a predicted test context; locating the group representation associated with the cognitive ability of the large language model through a parameter parsing technology, and improving the performance of the large language model on the specific task.

The application discloses a method and device for single-sample action recognition based on WiFi, which physically models the WiFi action recognition scene, generates virtual action data from basic actions, thereby enriching the data set and reducing the cost of collecting real data. Based on supervised learning and meta-learning, a single-sample action recognition framework is designed and implemented. The framework uses a traditional supervised learning mechanism for first-stage training on a virtual action data set, uses a single-sample meta-learning mechanism for second-stage training on a basic action data set, and uses a single-sample meta-learning mechanism for model fine-tuning on a single-sample data set of a new action, thereby obtaining a model capable of accurately recognizing the new action. The training of the two stages only needs to be performed at the initial deployment, and if there is a need to change the action type subsequently, only the model fine-tuning needs to be performed by using a single-sample data of a new action, thereby significantly reducing the model training cost, being highly scalable, and being applicable to actual application scenarios.

The present application belongs to the field of artificial intelligence, and relates to a text classification method, device, computer equipment and storage medium, the method comprising: segmenting a training text to obtain a plurality of semantic blocks; inputting each semantic block into an initial classification value assessment model to obtain a classification value score; screening the semantic blocks according to the classification value score to obtain a semantic block queue, and then inputting the initial text classification model to obtain a classification prediction result; determining the semantic block label of each semantic block in the semantic block queue; calculating a first loss based on the semantic block label and the classification value score, and calculating a second loss based on the classification prediction result and the classification label of the training text; adjusting the initial classification value assessment model according to the first loss to obtain a classification value assessment model, and adjusting the initial text classification model according to the second loss to obtain a text classification model; processing the text to be classified through the classification value assessment model and the text classification model to obtain a text classification result. The present application improves the accuracy of text classification.

The present invention relates to a remote sensing image cross-modal retrieval method based on the fusion of language and visual detail features, which includes the following steps: step 1: processing the training data of the remote sensing image-text retrieval model; step 2: constructing a multi-detail language and visual fusion model; step 3: training a multi-objective optimized detail language and visual fusion model; step 4: constructing a remote sensing image-text description feature library; step 5: completing cross-modal retrieval of remote sensing image-text description. The present invention uses a single-modal encoder to represent image and text features respectively, uses a multi-modal encoder to fuse the features of the two modalities, improves the expression of fine-grained semantic features of each encoder for the corresponding modal data through feature fusion and multi-task optimization training, and completes cross-modal retrieval through the similarity calculation of semantic features; by designing a multi-objective optimization strategy for the model, the model has the ability to express multiple details of features for remote sensing images and text descriptions at the same time.

The application relates to a training sample screening method and device based on global semantic modeling. The method comprises the following steps: obtaining at least one training sample set, each training sample set comprising at least one training sample; for each training sample set, inputting each training sample into an artificial intelligence model to obtain a response result of each training sample; the response result is obtained by at least performing self-attention calculation and nonlinear transformation calculation on the training sample by the artificial intelligence model; comparing whether the response result of each training sample meets corresponding labeled data of each training sample to determine a response accuracy of the artificial intelligence model on the training sample set; if the response accuracy meets a preset condition, the training sample set is determined as a target post-training sample set; and the target post-training sample set is used for performing post-training on the artificial intelligence model. By using the method, high-quality and low-cost post-training can be realized.

The invention provides a dynamic sign language recognition method based on AFFA and 3D FACNN-LSTM, and aims to solve the problems that in the prior art, due to the fact that a large amount of invalid and repeated data exists in preprocessing of sign language video data, the network training amount is large, and the dynamic gesture time sequence feature extraction effect is affected. The method mainly comprises the steps of preprocessing, AFFA duplicate removal, training test set division, attitude feature extraction and fusion, recognition and classification, model parameter adjustment, recognition and detection and the like, a spatio-temporal feature extraction method is used, and the network feature extraction effect is improved through a 3D feature attention module; through the self-adaptive frame taking algorithm, invalid and repeated data are effectively screened out, the training overhead is reduced, the network training effect is improved, the proposed network model parameter quantity is smaller than that of a traditional model, and the training overhead is lower.

The embodiment of the invention provides a multi-task image classification method and device, equipment and a storage medium, and relates to the technical field of image processing. The method comprises the following steps: performing pulse feature extraction on a sample image sequence by using a task encoder to obtain a task encoding vector; and obtaining model adjustment parameters based on the task coding vector by using a parameter generator, performing parameter adjustment on a basic pulse neural module in the basic network to obtain a modulation basic network, performing feature mapping on the sample image sequence by using the modulation basic network to obtain a sample embedding vector, and further obtaining a prototype-like vector. And determining a classification result of the to-be-processed image sequence according to the vector distance between the to-be-processed image sequence and each class prototype vector. Only one integral image classification model is needed to realize the classification process of image data of different tasks and different modalities. Because the time step length of each sample image sequence is basically consistent, the training time consumption is small, the training overhead is reduced, and the overall classification performance and efficiency are improved.

The invention discloses an image position identification method based on a twin width hash network. In the offline stage, firstly, a plurality of reference points are deployed in a target area, and 80%, 10% and 10% are taken to divide a training set, a verification set and a test set; then, constructing a Hash width network, and transversely increasing feature nodes through incremental learning so as to expand the width of the network; then, dividing positive and negative sample pairs, taking a width hash network as two branches of a twin network framework, and sharing weights and network parameters; and finally, keeping the similarity of the original samples to the greatest extent by shortening the Hash distance of similar features and shortening the Hash distance of dissimilar features. In the online stage, a query picture is input into the constructed model, a hash code with high semantic consistency with an original image is obtained, a most similar retrieval image is obtained through retrieval in a database, and the RP position corresponding to the retrieval image serves as a visual position recognition result. The invention provides an image position identification method based on a twin width Hash network by utilizing width learning transverse expansion nodes and fully considering intra-group correlation and inter-group exclusiveness, and the positioning precision is ensured while the system overhead is reduced.

The application discloses an adversarial knowledge distillation method and system based on feature semantic guidance and feature orthogonalization, which firstly acquires an image classification dataset, extracts text features, and constructs a semantic feature space.Secondly, a knowledge distillation framework including a teacher model, a student model and a projection head is constructed, the input of the teacher model and the student model is an image in the dataset, and the output is an image category prediction result. Then, the student model parameters are frozen, the projection head is updated through momentum tracking, in the feature level semantic adversarial sample generation stage, all network parameters are frozen, and normalized disturbance is constructed to obtain adversarial features. Finally, in the student model multi-objective joint optimization stage, the teacher model and the two projection heads are fixed, the student model parameters are updated, the classification loss, the distillation loss, the consistency loss and the orthogonalization loss are introduced, and optimization is completed. The application improves the fine-grained discrimination and generalization ability of the student model image classification, and does not introduce additional reasoning overhead.

The invention relates to the technical field of remote sensing image processing, in particular to a method, a system, equipment and a medium for generating remote sensing image changing subtitles, and the method comprises the following steps: acquiring a to-be-processed dual-time remote sensing image, and inputting the to-be-processed dual-time remote sensing image into a constructed changing subtitle generator to obtain dual-time image changing subtitles; wherein the changed subtitle generator comprises an image processing model and a changed subtitle generation model; the image processing model is obtained based on a neural network model and is used for processing the input dual-time remote sensing image to obtain a first feature, the first feature is input to the changed subtitle generation model, and the first feature is processed according to an instruction to generate changed subtitles corresponding to the dual-time image. According to the method, the details, the accuracy and the context correlation of the changed subtitles generated by the dual-time remote sensing image are remarkably improved.

The invention relates to the technical field of natural language processing, and discloses a large language model context learning improvement method based on parameter analysis, which comprises the following steps of: forming a verification prompt prefix by a task example sample and a verification preceding text through a template of a specific task, and inputting the verification prompt prefix into a large language model to obtain a prediction verification following text; performing group characterization division on the large language model, observing the performance reduction degree of the large language model on the specific task after the large language model is interfered through the evaluation indexes, and establishing association between the specific task and group characterization; the capability corresponding to group representation is improved through a representation improvement module; forming a test prompt prefix by the task example sample and the test preceding text through a template of a specific task, and inputting the test prompt prefix into a large language model in which a representation lifting module is introduced to obtain a prediction test following text; group characterization associated with the cognitive ability of the large language model is positioned through a parameter analysis technology, and the performance of the large language model on a specific task is improved.

The invention discloses a privacy protection image classification model training method and an image classification method and device, a privacy protection image classification model comprises a teacher image classification sub-model and a student image classification sub-model, and the image classification method comprises the following steps: obtaining to-be-classified image data; classifying and identifying the to-be-classified image through the privacy protection image classification model, and obtaining the classification of the to-be-classified image on the premise of privacy protection; the privacy protection image classification model is obtained through the training method of the privacy protection image classification model. The device comprises a privacy protection image classification model training device and an image classification device. According to the invention, classification and identification of the to-be-classified image are realized through the privacy protection image classification model.

The invention provides a model training method and device, and relates to the field of machine learning. The method and the device are used for improving model training efficiency. The method comprises: training a first model; a loss function value of the first model is monitored. And after determining that the loss function value of the first model meets a preset condition, amplifying the first model to obtain a second model. Wherein the amplified parameters in the second model are obtained by copying the parameters in the first model. And training the second model.

The present application relates to the technical field of computer vision, in particular to a pig part zero-label recognition method based on discrete Morse theory. A video frame sequence containing pigs is obtained, and instance segmentation of the pigs is performed based on a text prompt through a visual basic model to obtain a binary mask and image embedding features; an Euclidean distance transform terrain field is predicted according to the image embedding features and the background area is removed; the Euclidean distance transform terrain field is taken as a discrete Morse function to determine topological critical points and persistent homology values; an initial skeleton is obtained by skeleton thinning the binary mask, and the terminal branches are pruned according to the persistent homology values to obtain a pig topological skeleton; the head, tail and leg are identified according to the pig center, end points in the pig topological skeleton, head direction and pig body axis to obtain a pig part recognition result. The present application does not require manual key point labeling, can reduce deployment cost and improve cross-scene applicability.

The invention discloses a malicious software family classification method based on static multi-feature optimization and fusion. The method comprises the steps that malicious software image features are extracted based on OfficientNet-B3; performing static API sequence extraction by adopting IDA Pro and Radare2 respectively, and fusing extraction results of the IDA Pro and the Radare2 to obtain a fused static API sequence; extracting static statistical characteristics of malicious software; carrying out feature optimization on the fused static API sequence on the basis of a BERT-Tiny embedding mechanism; on the basis of the extracted and optimized features, calculating mutual information values of the features by adopting a discrete target variable mutual information algorithm, and screening out an optimal fusion feature sequence combination on the basis of the mutual information values of the features; and inputting the optimal fusion feature sequence combination into a classification model to complete malicious software family classification. The method solves the problems of single feature type, much redundant information, low classification efficiency and the like.