Search

Weakly supervised person search targets to detect and identify a person with only bounding box annotations. Recent approaches have focused on learning person relations in a single model, ignoring the conflicts between the detection and Re-ID heads, along with the influence of background elements, which may lead to noisy pseudo labels and inaccurate Re-ID features. To address this challenge, we introduce a novel framework named Knowledge Consistency Distillation (KCD) for weakly supervised person search, which explores the capabilities of an advanced unsupervised person re-identification (Re-ID) model to mitigate the conflicts and background influences. We propose hierarchical consistency alignments, including feature-level, cluster-level, and instance-level consistency alignment, to synchronize the knowledge from the state-of-the-art unsupervised Re-ID model. Specifically, the feature-level consistency aligns the feature through both context and relation alignment. The cluster-level consistency aligns the teacher cluster information by reusing its OIM module. To tackle the inconsistency problem between student instances and teacher cluster centroids, we incorporate pseudo-label refinement to assist the student model in comprehending the teacher’s knowledge at cluster-level while mitigating the negative effects of noisy labels. Finally, an instance-level consistency loss weighted by the similarity between the instance and its corresponding cluster is proposed to align the positive instance correlations. Our approach aims to train a one-step weakly supervised model for person search by exploiting the characteristics of unsupervised person Re-ID. Extensive experiments illustrate that our method achieves state-of-the-art performance on two widely-used person search datasets, CUHK-SYSU and PRW. Our code will be available on GitHub at https://github.com/zongyi1999/KCD.

Many existing graph-based methods for few-shot learning problem focused on either separately learning node features or edge features or simply utilizing graph convolution, failing to fully retain or exploit graph structure information. In this article, we proposed a bidirectional gated edge-labeling graph recurrent neural network (bi-GEGRN) which adopts both edge-labeling graph framework and graph convolution operation in the meta-learning scheme. We modified the gated graph neural network to adjacency matrix generator-based bidirectional formation which is able to process sequence graph data in two directions and then organically combined it with edge-labeled graph framework to cyclically upgrade features meanwhile aggregate graph structure information. In view of the excellent aggregating capability of graph convolution and good performance of the alternately cyclic update strategy, bi-GEGRN improves the information transferring between tasks in meta learning. To verify the validity and universality on both supervised and semi-supervised regimes, extensive experiments were conducted on three few-shot benchmark data sets and bi-GEGRN showed a good performance.

Few-shot learning aims at heuristically resolving new tasks with limited labeled data; most of the existing approaches are affected by knowledge learned from similar experiences. However, interclass barriers and new samples insufficiency limit the transfer of knowledge. In this article, we propose a novel mixture distribution graph network, in which the interclass relation is explicitly modeled and propagated via graph generation. Owing to the weighted distribution features based on the Gaussian mixture model, we take class diversity into consideration, thereby utilizing information precisely and efficiently. Equipped with minimal gated units, the “memory” of similar tasks can be preserved and reused through episode training, which fills a gap in temporal characteristics and softens the impact of data insufficiency. Extensive trials are carried out based on the MiniImageNet and CIFAR-FS data sets. The results turn out that our method exceeds most state-of-the-art approaches, which shows the validity and universality of our method in few-shot learning.

Adult image and video recognition is an important and challenging problem in the real world. Low-level feature cues do not produce good enough information, especially when the dataset is very large and has various data distributions. This issue raises a serious problem for conventional approaches. In this article, we tackle this problem by proposing a deep multicontext network with fine-to-coarse strategy for adult image and video recognition. We employ a deep convolution networks to model fusion features of sensitive objects in images. Global contexts and local contexts are both taken into consideration and are jointly modeled in a unified multicontext deep learning framework. To make the model more discriminative for diverse target objects, we investigate a novel hierarchical method, and a task-specific fine-to-coarse strategy is designed to make the multicontext modeling more suitable for adult object recognition. Furthermore, some recently proposed deep models are investigated. Our approach is extensively evaluated on four different datasets. One dataset is used for ablation experiments, whereas others are used for generalization experiments. Results show significant and consistent improvements over the state-of-the-art methods.

In existing local and global consistency (LGC) framework, the cost functions related to classifying functions adopt the sum of each row of weight matrix as an important factor. Some of these classifying functions are successfully applied to saliency detection. From the point of saliency detection, this factor is inversely proportional to the colour contrast between image regions and their surroundings. However, an image region that holds a big colour contrast against it surroundings does not denote it must be a salient region. Therefore a suitable variant of LGC is introduced by removing this factor in cost function, and a suitable classifying function (SCF) is decided. Then a saliency detection method that utilises the SCF, content-based initial label assignment scheme, and appearance-based label assignment scheme is presented. Via updating the content-based initial labels and appearance-based labels by the SCF, a coarse saliency map and several intermediate saliency maps are obtained. Furthermore, to enhance the detection accuracy, a novel optimisation function is presented to fuse the intermediate saliency maps that have a high detection performance for final saliency generation. Numerous experimental results demonstrate that the proposed method achieves competitive performance against some recent state-of-the-art algorithms for saliency detection.

AbstractCurrently, researches on content based image copy detection mainly focus on robust feature extraction. However, most of existing approaches use only a single feature to represent an image for copy detection, which is often insufficient to characterize the image content. Besides, with the exponential growth of online images, it's urgent to explore a way of tackling the problem of large scale. In this paper, we propose a feature fusion based hashing method which effectively utilize the correlation between two feature models and efficiently accomplish large scale image copy detection. To accurately map images into the Hamming space, our hashing method not only preserves the local structure of individual feature but also globally consider the local structures for all the features to learn a group of hash functions. The experiment results show that the proposed method outperforms the state-of-the-art techniques in both accuracy and efficiency.

It is a NP-hard problem to optimize the objective function of hash-based similarity search algorithms, such as Spectral Hashing and Self-Taught Hashing. To make the problem solvable, existing methods have relaxed the constraints on hash codes from binary values (discrete) to real values (continuous). Then eigenvalue decomposition was employed to achieve the relaxed real solution. The main problem is that the signs of the relaxed continuous solution are mixed. Such results may deviate severely from the true solution, which has lead to significant semantic loss. Moreover, eigenvalue decomposition confronts singularity problem when the dimension of the data is larger than the sample size. To address these problems, we propose a novel method named Nonnegative Sparse Locality Preserving Hashing (NSLPH). Nonnegative and sparse constraints are imposed for a more accurate solution which preserves semantic information well. Then, we have applied nonnegative quadratic programming and multiplicative updating to solve the optimization problem, which successfully avoids the singularity problem of the eigenvalue decomposition. The extensive experiments presented in this paper demonstrate that the proposed approach outperforms the state-of-the-art algorithms. [ABSTRACT FROM AUTHOR]

Abstract: The robustness of the localized watermarking methods mainly depends on the robustness of the feature locating the watermark. Based on the mean luminance of the disk, a rotation and scale invariant feature extraction algorithm is proposed. A theoretical verification of the rotation and scale invariance of extracted feature points in the continuous image is further performed. The extracted feature points are used to construct rotation and scale invariant circular regions, where the watermark is embedded after affine normalization. Experimental results show that the constructed regions fit the watermarking applications much better than those in previous feature-based watermarking schemes from the aspect of robustness against common attacks including filtering, JPEG compression, cropping, rotation and scaling, and the proposed localized image watermarking scheme has better robustness than previous feature-based watermarking schemes against common signal process and geometrical attacks while maintaining imperceptibility. [Copyright &y& Elsevier]

In this paper, the authors investigate the prospect of using multi-resolution histograms (MRH) in conjunction with digital image forensics, particularly in the detection of two kinds of copy-move manipulations, i.e., cloning and splicing. To the best of the authors’ knowledge, this is the first work that uses the same feature in both cloning and splicing forensics. The experimental results show the simplicity and efficiency of using MRH for the purpose of clone detection and splicing detection.