Images acquired under sand-dust weather conditions are severely degraded, with low contrast and severe color shift. The reason is that, due to the influence of sand-dust particles, light is scattered and absorbed, resulting in a blurred image and low contrast; the color shift is caused by the rapid attenuation of blue light. Therefore, to solve the problem of color shift and poor visibility in sand-dust images, this paper proposes a sand-dust image restoration method based on reversing the blue channel prior (RBCP). Under the influence of the blue channel, the dark channel prior (DCP) method will fail. Therefore, the method first reverses the blue channel of the sand-dust image and uses the dark channel prior method, which we call RBCP, and then, RBCP is used to estimate the atmospheric light and transmission map and recover the sand-dust image. The restored image shows significantly improved visibility. When estimating the transmission map, a guiding filter is used to improve the coarse transmission map, and a tolerance mechanism is introduced to modify the transmission map of bright areas in the sky to solve the problem of distortion in the sky. Finally, combined with the gray world, an adaptive color adjustment factor is introduced into the restoration model to remove the color shift. Experimental results via qualitative and quantitative evaluation demonstrate that the proposed method can effectively recover clear sand-dust images and produce results superior to those of other state-of-the-art methods.

No abstract

To address the problem that some current algorithms suffer from the loss of some important features due to rough feature distillation and the loss of key information in some channels due to compressed channel attention in the network, we propose a progressive multistage distillation network that gradually refines the features in stages to obtain the maximum amount of key feature information in them. In addition, to maximize the network performance, we propose a weight-sharing information lossless attention block to enhance the channel characteristics through a weight-sharing auxiliary path and, at the same time, use convolution layers to model the interchannel dependencies without compression, effectively avoiding the previous problem of information loss in channel attention. Extensive experiments on several benchmark data sets show that the algorithm in this paper achieves a good balance between network performance, the number of parameters, and computational complexity and achieves highly competitive performance in both objective metrics and subjective vision, which indicates the advantages of this paper’s algorithm for image reconstruction. It can be seen that this gradual feature distillation from coarse to fine is effective in improving network performance. Our code is available at the following link: https://github.com/Cai631/PMDN.

No abstract

Violence detection aims to locate violent content in video frames. Improving the accuracy of violence detection is of great importance for security. However, the current methods do not make full use of the multi-modal vision and audio information, which affects the accuracy of violence detection. We found that the violence detection accuracy of different kinds of videos is related to the change of optical flow. With this in mind, we propose an optical flow-aware-based multi-modal fusion network (OAMFN) for violence detection. Specifically, we use three different fusion strategies to fully integrate multi-modal features. First, the main branch concatenates RGB features and audio features and the optical flow branch concatenates optical flow features with RGB features and audio features, respectively. Then, the cross-modal information fusion module integrates the features of different combinations and applies weights to them to capture cross-modal information in audio and video. After that, the channel attention module extracts valuable information by weighting the integration features. Furthermore, an optical flow-aware-based score fusion strategy is introduced to fuse features of different modalities from two branches. Compared with methods on the XD-Violence dataset, our multi-modal fusion network yields APs that are 83.09% and 1.4% higher than those of the state-of-the-art methods in offline detection, and 78.09% and 4.42% higher than those of the state-of-the-art methods in online detection.