Traffic sign detection aims to locate and classify traffic signs in real time and accurately. But because of their small size and complex backgrounds, some smaller traffic signs are harder to detect than larger ones. On the other hand, some false information is always detected due to the influence of light changes and bad weather. Therefore, in order to solve the problems of missing detection and false detection, this paper proposes an adaptive multi-scale spatial-channel attention fusion center network (MSCA-Center Net). Firstly, a residual space-channel attention module combined with multi-channel information is proposed. The module divides the channels in the feature map into several groups, generate separate spatial and channel attention for each group, and combine the spatial and channel information of the multi-channel. Then, a multi-scale attention fusion module is proposed to integrate the extracted high-level and low-level features, which can improve the detection and classification accuracy. In addition, a coordinate attention module is introduced to obtain the final feature code, so that the model can locate and identify the target area more accurately. Finally, the optimal detection box is generated adaptively according to the feature coding. The proposed model was tested and evaluated on the CCTSDB dataset. Compared with the existing methods, the proposed method can detect traffic signs adaptively in real time under complex background, which verifies the effectiveness of the proposed method.

Facial expression recognition(FER) is one of the important research in computer vision, which has been widely applied in human-computer interaction, education, healthcare, transportation, etc. However, the wide application of facial expression recognition technology also brings new challenges, where occlusion and pose variation are two of the worst factors that disturb facial expression recognition in the wild. We propose a facial expression recognition method based on local feature augmentation and multi-scale global correlation which can adaptively extract robust local features and global features from the feature level to suppress the disturbances of occlusion and pose variation on facial expression recognition. The experimental results show that our method performs well on the RAF-DB dataset and has stronger robustness compared with other algorithms.

Abstract With the development of image recognition technology, face, body shape, and other factors have been widely used as identification labels, which provide a lot of convenience for our daily life. However, image recognition has much higher requirements for image conditions than traditional identification methods like a password. Therefore, image enhancement plays an important role in the process of image analysis for images with noise, among which the image of low-light is the top priority of our research. In this paper, a low-light image enhancement method based on the enhanced network module optimized Generative Adversarial Networks(GAN) is proposed. The proposed method first applied the enhancement network to input the image into the generator to generate a similar image in the new space, Then constructed a loss function and minimized it to train the discriminator, which is used to compare the image generated by the generator with the real image. We implemented the proposed method on two image datasets (DPED, LOL), and compared it with both the traditional image enhancement method and the deep learning approach. Experiments showed that our proposed network enhanced images have higher PNSR and SSIM, the overall perception of relatively good quality, demonstrating the effectiveness of the method in the aspect of low illumination image enhancement.

Data deduplication, though being efficient in removing duplicate chunks, introduces chunk fragmentation which decreases restore performance. Rewriting algorithms are proposed to reduce the chunk fragmentation. Delta compression is often used as a complement for data deduplication to further improve storage efficiency. We observe that delta compression introduces a new type of chunk fragmentation stemming from improper delta compression for chunks of which the base chunks are fragmented. The new type of chunk fragmentation severely decreases restore performance and cannot be addressed by existing rewriting algorithms. To address this problem, we propose SDC, a scheme performing post-deduplication delta compression only for the chunks of which the bases can be directly found in the restore cache to eliminate additional disk reads for base chunks, thus avoiding the new type of chunk fragmentation. In addition, self-referenced chunks can be fragmented, which decrease restore performance, and these fragmented chunks can serve as bases to decrease the restore performance repeatedly. We propose a hybrid rewriting scheme for SDC to rewrite such fragmented chunks. Experimental results show that SDC improves the restore performance of the approach that directly performs delta compression after data deduplication by 2.9-16.9x, and achieves more than 95 percent of its compression gains.

Hashing algorithm has been widely used to speed up image retrieval due to its compact binary code and fast distance calculation. The combination with deep learning boosts the performance of hashing by learning accurate representations and complicated hashing functions. So far, the most striking success in deep hashing have mostly involved discriminative models, which require labels. To apply deep hashing on datasets without labels, we propose a deep self-taught hashing algorithm (DSTH), which generates a set of pseudo labels by analyzing the data itself, and then learns the hash functions for novel data using discriminative deep models. Furthermore, we generalize DSTH to support both supervised and unsupervised cases by adaptively incorporating label information. We use two different deep learning framework to train the hash functions to deal with out-of-sample problem and reduce the time complexity without loss of accuracy. We have conducted extensive experiments to investigate different settings of DSTH, and compared it with state-of-the-art counterparts in six publicly available datasets. The experimental results show that DSTH outperforms the others in all datasets.

Data deduplication, a data redundancy elimination technique, has been employed in almost all kinds of application environments to reduce storage space. However, one of the main challenges facing deduplication technology is to provide a fast key-value fingerprint index for large datasets, as the index performance is critical to the overall deduplication performance. This paper proposes RMD, a resemblance and mergence based deduplication scheme, which aims to provide quick responses to fingerprint queries. The key idea of RMD is to leverage a bloom filter array and the data resemblance algorithm to dramatically reduce the query range for deduplication. Moreover, RMD utilizes mergence based approach to merge resemblance segments to relevant bins, and exploits frequency-based Fingerprint Retention Policy to reduce the bin capacity to improve query throughput and improve data deduplication ratio. Extensive experimental results with real-world datasets have shown that RMD is able to achieve pretty high query performance and outperforms several state-of-the-art deduplication schemes.

Hashing algorithm has been widely used to speed up image retrieval due to its compact binary code and fast distance calculation. The combination with deep learning boosts the performance of hashing by learning accurate representations and complicated hashing functions. So far, the most striking success in deep hashing have mostly involved discriminative models, which require labels. To apply deep hashing on datasets without labels, we propose a deep self-taught hashing algorithm (DSTH), which generates a set of pseudo labels by analyzing the data itself, and then learns the hash functions for novel data using discriminative deep models. Furthermore, we generalize DSTH to support both supervised and unsupervised cases by adaptively incorporating label information. We use two different deep learning framework to train the hash functions to deal with out-of-sample problem and reduce the time complexity without loss of accuracy. We have conducted extensive experiments to investigate different settings of DSTH, and compared it with state-of-the-art counterparts in six publicly available datasets. The experimental results show that DSTH outperforms the others in all datasets.

With the explosive increase of online images, fast similarity search is increasingly critical for large scale image retrieval. Several hashing methods have been proposed to accelerate image retrieval, a promising way is semantic hashing which designs compact binary codes for a large number of images so that semantically similar images are mapped to similar codes. Supervised methods can handle such semantic similarity but they are prone to overfitting when the labeled data is few or noisy. In this paper, we concentrate on this issue and propose a novel Inductive Transfer Deep Hashing (ITDH) approach for semantic hashing based image retrieval. A transfer deep learning algorithm has been employed to learn the robust image representation, and the neighborhood-structure preserved method has been used to mapped the image into discriminative hash codes in hamming space. The combination of the two techniques ensures that we obtain a good feature representation and a fast query speed without depending on large amounts of labeled data. Experimental results demonstrate that the proposed approach is superior to some state-of-the-art methods.

Digital fingerprinting is a promising approach to protect multimedia content from unauthorized redistribution. However, the existing fingerprints are unsuitable for social network tasks, because they fail to represent the social network structure, which incurs inefficient fingerprint coding. In addition, they are infeasible to efficiently trace colluders due to the large scale of social networks. To address these problems, we design a novel fingerprint, which consists of community relationship code and user identification code. Aiming to preserving the social network structure, we propose a kernelized neighborhood preserving hashing method to generate community relationship codes. The proposed method assigns similar community relationship codes to users in the same or close communities, which improves the anticollusion performance. Because the community relationship codes are binary and neighborhood preserving, they can be used for fast indexing and retrieval. To accelerate the collusion fingerprint tracing, we treat the community relationship codes as index keys to construct a hash table and an inverted index table. Based on the tables, we correspondingly propose an efficient fingerprint detection method. Extensive experiments show that the proposed fingerprint is suitable for social network tasks and the real colluders can be efficiently identified by the proposed fingerprint detection approach.

Fast similarity search methods are increasingly critical for many large-scale learning tasks, particularly in the communities of machine learning and data mining. Recently, data-aware hashing method is regarded as a promising approach for similarity search which maps high-dimensional feature vectors into efficient and compact hash codes while preserving the corresponding neighborhood structure. Although some recent hashing methods based on eigenvalue decomposition perform well, they suffer from semantic loss. In this paper, we concentrate on this issue and propose a novel neighborhood preserving hashing approach which adopts a brand-new method to combine non-negative matrix factorization and locality linear embedding without introducing any additional parameter. The combination of these two classical techniques ensures that we obtain a parts-based representation which not only fulfill the psychological and physiological requirements of human perception but also conserve the intrinsic neighborhood structure of the original data. Experiments are conducted to demonstrate that the proposed approach is superior to some state-of-the-art methods.