Attention-based hybrid contrastive learning for unsupervised person re-identification

Person re-identification (ReID) aims at searching the same identity person among images captured by various cameras. Existing fully supervised person ReID methods usually suffer from poor generalization capability caused by domain gaps. Unsupervised person ReID has attracted a lot of attention recently, because it works without intensive manual annotation and thus shows great potential in adapting to new conditions. Representation learning plays a critical role in unsupervised person ReID. In this work, we propose a novel selective contrastive learning framework for fully unsupervised feature learning. Specifically, different from traditional contrastive learning strategies, we propose to use multiple positives and adaptively selected negatives for defining the contrastive loss, enabling to learn a feature embedding model with stronger identity discriminative representation. Moreover, we propose to jointly leverage global and local features to construct three dynamic memory banks, among which the global and local ones are used for pairwise similarity computation and the mixture memory bank are used for contrastive loss definition. Experimental results demonstrate the superiority of our method in unsupervised person ReID compared with the state of the art. Our code is available at https://github.com/pangbo1997/Unsup_ReID.git .

Efficient attention based deep fusion CNN for smoke detection in fog environment

Older People Issues and Innovations in Care 3rd edition

Currently, the mainstream unsupervised method uses a clustering algorithm to cluster unlabeled data and generate labels for training samples based on the clustering results. The critical step is to obtain the features. Most unsupervised methods are based on single-branch networks. Still, multi-branch networks that fusion local and global features have been proved to improve supervised effectively. This paper studies the problem of unsupervised person re-identification using the multi-branch network. The goal is to extract more reliable feature representation through the multi-branch network and obtain a model with a more vital distinguishing ability. Specifically, an OSNet-based multi-branch network simultaneously extracts global, local, and channel features. It applies a bottom-up clustering algorithm based on the hierarchy to create labels for unlabeled training samples and uses Softmax-Triplet joint loss to optimize the model. We verify the proposed method on two benchmark re-ID datasets, such as the Market-1501 and DukeMTMC-reID datasets. Compared with the baseline method, the accuracy of Rank-1 and mAP are improved by 5.5 and 12.6% on Market-1501, and 5.5 and 7.6% on DukeMTMC-reID.

Attention mechanism-based deep learning has played an important role in vision-based smoke detection in forest fire early warning systems. However, the lack of consideration of the specific characteristics of smoke may render existing attention mechanisms ineffective, e.g., in detecting small areas of smoke, particularly in the complex forest smoke scenes involving multiple points of ignition or varying scales. To enhance the accuracy and the interpretability of smoke detection, we propose a Spatial and Efficient Channel Attention mechanism, termed SECA, and integrate SECA into deep models to incorporate the characteristics of smoke diffusion. Technically, multi-kernel 1-dimensional (1D) convolution is utilized for multi-scale smoke-capturing, to replace single-kernel 2D or 3D convolution in existing methods. In implementation, our SECA mechanism can also be used as a common module and easily plugged into a backbone network. To accelerate our model, a DSConv-Haar Wavelet Downsampling technique called DHWD is also provided. Extensive experiments were conducted on public datasets and self-collected datasets. Compared to existing methods, our method can achieve a better or at least a comparable performance in smoke detection in terms of smoke detection accuracy, computational efficiency, and ease of use. For example, it surpasses baseline methods, demonstrating average improvements of 4.2% in mAP50 and of 3.7% in mAP50-95, respectively.

Traditional person re-identification methods are mostly based on supervised state of manually annotated data sets. Although the recognition accuracy is high, the algorithm completely relies on effective annotation information and lacks practicability. Based on the above problems, an unsupervised person re-identification method is proposed, which refines features to guide multi-label distribution sorting learning. Firstly, a single and multi-label classification network structure is constructed to improve the matching accuracy of tags by learning the characteristics of each person and its k nearest neighbor and sharing the same class. Secondly, the refinement feature extraction module is designed and embedded into the original ResNet50 network, and the potential key areas in the person image are adaptively located. While weakening the appearance differences of the features, multi-information mining is carried out to obtain the refined features of the characters and assist the classification network to improve the performance. Thirdly, reliable multi-category labels are screened by punishing the sorting between positive and negative categories of labels in multi-category labels. Finally, the network is supervised by combining multi-label distribution sorting loss and multi-label classification loss. Experiments show that without any annotation information, the recognition accuracy of Market-1501 and DukeMTMC-reID datasets is 58.4% and 42.3%, respectively, and the mAP of Market-1501 dataset is increased by 12.9%.

Learning discriminative image representations plays a vital role in long-tailed image classification because it can ease the classifier learning in imbalanced cases. Given the promising performance contrastive learning has shown recently in representation learning, in this work, we explore effective supervised contrastive learning strategies and tailor them to learn better image representations from imbalanced data in order to boost the classification accuracy thereon. Specifically, we propose a novel hybrid network structure being composed of a supervised contrastive loss to learn image representations and a cross-entropy loss to learn classifiers, where the learning is progressively transited from feature learning to the classifier learning to embody the idea that better features make better classifiers. We explore two variants of contrastive loss for feature learning, which vary in the forms but share a common idea of pulling the samples from the same class together in the normalized embedding space and pushing the samples from different classes apart. One of them is the recently proposed supervised contrastive (SC) loss, which is designed on top of the state-of-the-art unsupervised contrastive loss by incorporating positive samples from the same class. The other is a prototypical supervised contrastive (PSC) learning strategy which addresses the intensive memory consumption in standard SC loss and thus shows more promise under limited memory budget. Extensive experiments on three long-tailed classification datasets demonstrate the advantage of the proposed contrastive learning based hybrid networks in long-tailed classification.

Most of existing unsupervised person re-identification methods focus on cross-domain adaptation. In order to further relieve the dependence on manual labels, we propose a camera-aware three-stage method for fully unsupervised person re-identification which only requires the unlabeled target dataset. We exploit camera labels and divide the learning process into three relatively easy sub-tasks: initialization by instance discrimination, intra-camera learning and inter-camera learning. The first stage regards each person image as an instance and tries to distinguish each image. The second stage performs intra-camera clustering while the last stage performs clustering and training on the whole dataset. These three stages share the backbone network. Finally, our method substantially boosts the performance stage by stage without any manual ID annotation. We conduct extensive experiments on three large-scale image-based datasets, including Market-1501, DukeMTMC-reID and MSMT17. The results demonstrate that our method achieves the state-of-the-art performance.

In view of the advantages of deep learning model in the extraction of abstract concept, a new text clustering algorithm is designed based on Deep Boltzmann Machines. Based on Replicate Softmax Model and new Deep Boltzmann Machine, energy function of this model is proposed and the detail learning algorithm is introduced.

Feature relocation network for fine-grained image classification

Case instance segmentation of small farmland based on Mask R-CNN of feature pyramid network with double attention mechanism in high resolution satellite images

Recently, the attention mechanism combining spatial and channel information has been widely used in various deep convolutional neural networks (CNNs), proving its great potential in improving model performance. However, this usually uses 2D global pooling operations to compress spatial information or scaling methods to reduce the computational overhead in channel attention. These methods will result in severe information loss. Therefore, we propose a Spatial channel attention mechanism that captures cross-dimensional interaction, which does not involve dimensionality reduction and brings significant performance improvement with negligible computational overhead. The proposed attention mechanism can be seamlessly integrated into any convolutional neural network since it is a lightweight general module. Our method achieves a performance improvement of 2.08% on ResNet and 1.02% on MobileNetV2 in top-one error rate on the ImageNet dataset.

Currently, breast cancer is one of the most common cancers among women. To aid clinicians in diagnosis, lesion regions in mammography pictures can be segmented using an artificial intelligence system. This has significant clinical implications. Clustering algorithms, as unsupervised models, are widely used in medical image segmentation. However, due to the different sizes and shapes of lesions in mammography images and the low contrast between lesion areas and the surrounding pixels, it is difficult to use traditional unsupervised clustering methods for image segmentation. In this study, we try to apply the semisupervised fuzzy clustering algorithm to lesion segmentation in mammography molybdenum target images and propose semisupervised fuzzy clustering based on the cluster centres of labelled samples (called SFCM_V, where V stands for cluster centre). The algorithm refers to the cluster centre of the labelled sample dataset during the clustering process and uses the information of the labelled samples to guide the unlabelled samples during clustering to improve the clustering performance. We compare the SFCM_V algorithm with the current popular semisupervised clustering algorithm and an unsupervised clustering algorithm and perform experiments on real patient mammogram images using DICE and IoU as evaluation metrics; SFCM_V has the highest evaluation metric coefficient. Experiments demonstrate that SFCM_V has higher segmentation accuracy not only for larger lesion regions, such as tumours, but also for smaller lesion regions, such as calcified spots, compared with existing clustering algorithms.

The traditional K-means clustering algorithm is difficult to determine the cluster number, which is sensitive to the initialization of the clustering center and easy to fall into local optimum. This paper proposes a clustering algorithm based on self-organizing mapping network and weight particle swarm optimization SOM&WPSO (Self-Organization Map and Weight Particle Swarm Optimization). Firstly, the algorithm takes the competitive learning mechanism of a self-organizing mapping network to divide the data samples into coarse clusters and obtain the clustering center. Then, the obtained clustering center is used as the initialization parameter of the weight particle swarm optimization algorithm. The particle position of the WPSO algorithm is determined by the traditional clustering center is improved to the sample weight, and the cluster center is the “food” of the particle group. Each particle moves toward the nearest cluster center. Each iteration optimizes the particle position and velocity and uses K-means and K-medoids recalculates cluster centers and cluster partitions until the end of the algorithm convergence iteration. After a lot of experimental analysis on the commonly used UCI data set, this paper not only solves the shortcomings of K-means clustering algorithm, the problem of dependence of the initial clustering center, and improves the accuracy of clustering, but also avoids falling into the local optimum. The algorithm has good global convergence.

Learning discriminative features between abnormal and normal instances is crucial for video anomaly detection within the multiple instance learning framework. Existing methods primarily focus on instances with the highest anomaly scores, neglecting the identification and differentiation of hard samples, leading to misjudgments and high false alarm rates. To address these challenges, we propose a dual triplet contrastive loss strategy. This approach employs dual memory units to extract four key feature categories: hard samples, negative samples, positive samples, and anchor samples. Contrastive loss is utilized to constrain the distance between hard samples and other samples, enabling accurate identification of hard samples and enhancing the discriminative ability of hard samples and abnormal features. Additionally, a multi-scale feature perception module is designed to capture feature information at different levels, while an adaptive global-local feature fusion module constructs complementary feature enhancement through feature fusion. Experimental results demonstrate the effectiveness of our method, achieving AUC scores of 87.16% on the UCF-Crime dataset and AP scores of 83.47% on the XD-Violence dataset.