34results about How to "Improve re-identification performance" patented technology

The invention relates to a pedestrian re-identification method based on a depth multi-loss fusion model. According to the method, a deep learning technology is used, pre-processing operations such asoverturning, cutting, random erasing and style migration are carried out on the training set picture. Feature extraction is carried out through a basic network model. A plurality of loss functions areused for carrying out fusion joint training on the network. Compared with a pedestrian re-identification algorithm based on deep learning, since the method uses a plurality of preprocessing modes, fusion of three loss functions and an effective training strategy, the pedestrian re-identification performance on the data set is greatly improved. On one hand, multiple preprocessing modes can expanda data set, the generalization capability of the model is improved, the occurrence of an overfitting condition is avoided, and on the other hand, the three loss functions have respective advantages and disadvantages, and when the three loss functions are effectively combined, the used model can obtain a better recognition result.

The invention discloses a pedestrian re-identifying method based on coordination scale learning and belongs to the technical field of monitoring video retrieval. First, according to color and texture features of images in a marked training sample set L, scale learning is carried out, and covariance matrixes Mc and Mt in corresponding Mahalanobis distance are obtained; and checking targets are selected randomly, the Mc and the Mt are used for Mahalanobis distance measuring, a corresponding sorting result is obtained, positive samples and negative samples are obtained, a new marked training sample set L is obtained, the Mc and the Mt are updated until an unmarked training sample set U is empty, a final marked sample set L* is obtained, the color and texture features are fused, an Mf is obtained, and a Mahalanobis distance function based on the Mf can be used for pedestrian re-identifying. Under a semi-supervised framework, the pedestrian re-identifying technology based on scale learning is studied, scale learning is carried out with the marked samples assisted by the unmarked samples, the requirement that practical video investigation application marked training samples are hard to obtain is met, and re-identifying performance under few marked samples can be effectively improved.

The invention discloses a multi-scale pedestrian re-identification method based on multi-granularity depth feature fusion, and the method comprises the steps: selecting a pedestrian re-identification data set, and carrying out the preprocessing of a training set in the data set; selecting a residual network as a basic skeleton, including a global coarse-grained fusion learning branch, a local coarse-grained fusion learning branch and a local attention fine-grained fusion learning branch; adopting softmax loss and triple loss as re-identification network supervisors, and training a pedestrian re-identification network model; and fusing the network features of different branches as a final descriptor of the pedestrian, and taking the to-be-queried pedestrian image as the input of the pedestrian re-identification network model to obtain a pedestrian re-identification result. According to the method, the pressure of a complex background or posture change on a re-identification task is effectively relieved, and the identification precision is improved.

The invention discloses a pedestrian re-recognition method based on similarity and dissimilarity fusion ranking optimization. Pedestrian re-recognition is a specific pedestrian retrieval problem. The main ideal of the pedestrian re-recognition method based on the similarity and dissimilarity fusion ranking optimization is that correctly-retrieved pedestrian targets are similar to polar positive samples of queried pedestrians and are not similar to polar negative samples of the queried pedestrians, meanwhile the pedestrians similar to the polar positive samples are defined as suspected positive samples, the samples similar to the polar negative samples defined as suspected negative samples, and the pedestrian re-recognition improving effect is achieved by improving ranking of suspected positive samples and meanwhile reducing ranking of suspected negative samples in a ranking optimization mode. In addition, the similarity and dissimilarity relation is strengthened by fusing results of multiple methods. By means of the pedestrian re-recognition method based on the similarity and dissimilarity fusion ranking optimization, similarity and dissimilarity ranking results are fused to further improve the matching accuracy of the same row of pedestrians under multiple cameras.

The invention discloses a pedestrian re-identification method based on a triple convolutional neural network. The method comprises the steps of 1, forming a small batch of samples; 2, outputting global features and local features of each pedestrian image; 3, calculating the distance between the global features of every two pedestrian images in the small-batch samples and the distance between the local features of the corresponding body parts; 4, forming P x K pairs of triple samples; 5, using a Log-logistic function to replace a hinge function to improve a triple convolutional neural network loss function; 6, obtaining a trained triple convolutional neural network; asnd step 7, identifying the target pedestrian image by using the trained triple convolutional neural network, and further avoiding mismatching between the target pedestrian image and the pedestrian image to be identified by combining a confidence measurement mechanism. According to the pedestrian re-identification method based on the triple convolutional neural network, pedestrian re-identification can be accurately carried out, and mismatching between different pedestrians is avoided.

The invention discloses a vehicle re-identification method based on a multi-loss fusion model. According to the method, a deep convolutional neural network structure is designed for the problem of vehicle re-identification, a multi-loss fusion model is adopted to jointly supervise the training of the deep convolutional neural network, the joint optimization of the same ID sample difference and different ID sample differences of vehicles is realized, and the purpose of learning feature expression with higher discriminability is achieved. Wherein the proposed multi-cluster center loss function can increase the inter-class distance and shorten the intra-class distance, so that the vehicle features belonging to the same ID are close to the class center as much as possible, and the discrimination of feature expression is effectively improved. According to the multi-loss fusion model provided by the invention, multiple data enhancement modes are combined, so that the vehicle re-identification precision is effectively improved.

The invention discloses a dynamic low-resolution pedestrian re-identification method based on scale distance gradient function interface learning, and is to obtain a good re-identification effect when images are low in resolution and change in scale. The method comprises the following steps: to begin with, obtaining an image pair, that is, a positive sample pair, from the same person, and image pairs, that is, negative sample pairs, from different persons; then, introducing distance scale gradient functions, wherein the positive sample pair generates a feasible distance scale gradient function, and the negative sample pair generates an infeasible distance scale gradient function, and the two kinds of distance scale gradient functions form a distance scale gradient function space, a positive parameter vector expressing the feasible distance scale gradient function, and a negative parameter vector expressing the infeasible distance scale gradient function; and finally, carrying out classification through a trained random forest classifier to judge whether the images are from the same object according to the information that whether some distance scale gradient function is in a positive domain or a negative domain in the function space.

The invention provides a pedestrian re-recognition method based on multi-task learning. The method comprises the steps: obtaining a cross-camera pedestrian image, and constructing a pedestrian re-recognition training data set; constructing a multi-task learning network, wherein the network can combine an attribute task and an identity recognition task, all parameters and losses can be optimized ina combined mode, and therefore the purpose of improving pedestrian re-recognition accuracy is achieved. The attribute task and the recognition task respectively comprise a verification step, a verification step, a classification step and a verification loss to optimize the distance of the sample, the recognition loss is used for constructing a large class space and verifying the loss at the sametime, and the space is optimized by minimizing similar images and maximizing the distance between different images so as to improve the recognition precision.

The invention discloses a semi-supervised character re-identification method based on camera styles and human body posture adaptation. The method comprises the following steps: selecting labeled datacovering all identities and all camera styles; carrying out data expansion on the labeled data to generate images with different postures in the camera and images with different styles between the cameras; training a network model by combining the expanded labeled data with the unlabeled data; and performing pseudo label distribution by a strategy of constraining the average distance between the label-free data and the label-containing data characteristics and the number of each type of images. According to the method, the problems of small label data volume and unreasonable pseudo label distribution in the prior art are solved.

The invention discloses a pedestrian re-identification method based on label uncertainty and a human body component model. The pedestrian re-identification method comprises the following steps: (1) constructing a deep neural network model based on a human body component; (2) initializing the constructed deep neural network model, and performing training according to the constructed deep neural network structure to obtain another model; (3) training the initialized deep neural network model and updating parameters in the network; (4) carrying out feature extraction on the target pedestrian image and pedestrian images in a pedestrian image library by adopting the trained deep neural network; And (5) performing cosine similarity calculation and sorting on the extracted features to obtain a recognition result. According to the method, the problems that local component information is not completely accurate and classification confidence coefficients of all local components are large in difference are effectively processed, the correct rate of pedestrian re-identification can be effectively improved, and/or the false identification rate can be effectively reduced.

The invention provides a pedestrian image recognition method based on a deep network model. The pedestrian image recognition method comprises the steps of performing data preprocessing on a pedestrianimage; executing an adaptive sampling algorithm on the preprocessed data to obtain a batch with more difficult samples; extracting multilayer features through a backbone network model; enhancing thelow-level features by using a sub-module, performing downscaling, splicing the low-level features with the high-level features to obtain multi-level features, segmenting the multi-level features according to different granularities to form a multi-branch structure, extracting component features and global features of each branch, and splicing all the extracted features to obtain depth representation of the pedestrian image; training the constructed network model; extracting the depth representation of the query images through the trained network model, and returning the recognition result of each query image according to the cosine distance similarity between each query image and the queried set. Through the multi-level multi-granularity pedestrian re-recognition depth model, the optimal pedestrian re-recognition performance at the present stage is realized.

The invention discloses a method for generating pedestrian re-identification data under different illumination conditions based on an adversarial network. The method comprises the following steps: firstly, acquiring videos of different time periods in a monitoring video, namely monitoring videos under different illumination conditions; Extracting representative monitoring images from the acquiredmonitoring videos, and marking the representative monitoring images into three different categories of'too dark, normal and too bright 'according to illumination conditions; Sending the images with the illumination condition labeling information into a generative adversarial network for training, wherein the trained generative network is used for generating data under different illumination conditions; And generating the pedestrian re-identification data into data under different illumination conditions through the trained generation network; And finally, adding the obtained data as a data augmentation form into pedestrian re-identification model training, and at the moment, enabling the model to identify the characteristics of the same target under different illumination conditions, i.e.,enhancing the identification performance of the model on the target under different illumination conditions.

The invention provides a cross-domain pedestrian re-identification method and system based on multi-feature mixed learning, and belongs to the technical field of computer vision. The method comprises: by means of a re-identification model subjected to combined training, extracting pedestrian global features, pedestrian attribute features and pedestrian local features of the pedestrian image to be recognized and a bottom library image, which is similar to the pedestrian identity in the pedestrian image to be recognized, in the image bottom library gallley; and fusing the extracted to-be-identified features, and carrying out similarity matching sorting on the fused features of the features of the bottom library image to obtain a pedestrian re-identification result. According to the method, inter-domain joint training and multi-feature mixed learning are utilized to reduce inter-domain differences, so that the system is more stable and higher in robustness, source domain training of global and local features and joint training of attribute features are performed on images of different scenes, pedestrian attributes are combined, the adaptive capacity of a cross-domain pedestrian re-identification model is improved, and pedestrian re-identification is carried out on a cross-domain data set, so that the cross-domain pedestrian re-identification performance is improved.

The invention belongs to the technical field of character recognition, and discloses a video-based re-identification method and system for people in a smoke scene and a terminal; and the method comprises the steps: constructing a symmetric non-local coding and decoding K estimation network model to carry out defogging of a video; constructing a discrimination network model, and estimating whetherthe input video is a normal video or a fog-free video generated by a defogging sub-network based on the constructed discrimination network model; and constructing a non-local double-attention character re-identification sub-network model, and re-identifying the character. According to the invention, the problem of difficult re-identification caused by foggy re-identification of people in the videocan be solved. According to the invention, people re-identification can be well completed in a foggy video. The whole process of the method is an end-to-end design, and the method can be used more simply. The method is a technology for re-identifying the person in the smoke scene based on the video, end-to-end judgment can be completed, and re-identification of the person can be well completed.

The invention discloses a pedestrian re-identification method for solving part misalignment, and the method comprises the steps: carrying out the data preprocessing of a pedestrian image, adjusting the size of the pedestrian image, carrying out the data enhancement, and carrying out the data standardization processing; constructing a pedestrian re-identification network model, namely constructingdeep representation of a pedestrian image, extracting multilayer features through a backbone network model, enhancing and fusing the multilayer features by using sub-modules to form a multi-branch structure, and extracting component features and global features of each branch; training the constructed network model, defining experiment-related configuration, and optimizing model parameters of thenetwork model; and pedestrian re-identification: extracting the depth representation of the query image through the trained network model, and returning the identification result of each query image according to the similarity between each query image and the queried set after using the second normal form normalization. The pedestrian re-identification method based on fusion of multi-scale features to solve the problem of component misalignment realizes the optimal pedestrian re-identification performance at the present stage.