A5100734549- Advanced Image Processing Techniques
- Image and Video Quality Assessment
- Sparse and Compressive Sensing Techniques
- Advanced Vision and Imaging
- Image Processing Techniques and Applications
- Image and Video Stabilization
- Advanced Image Fusion Techniques
- Image and Signal Denoising Methods
- Advanced Technologies in Various Fields
University of Electronic Science and Technology of China
2023-2024
Video super-resolution (VSR) is used to compose high-resolution (HR) video from low-resolution video. Recently, the deformable alignment-based VSR methods are becoming increasingly popular. In these methods, features extracted aligned eliminate motion error targeting high (SR) quality. However, often suffer misalignment and lack of enough temporal information HR frames, which accordingly induce artifacts in SR result. this article, we design a deep network (DVSRNet) based on proposed...
This paper presents a Bidirectional Feature Fusion Network (BiDFNet) for 3D object detection, leveraging pseudo-point clouds to achieve bidirectional fusion of point cloud and image features. The proposed model addresses key challenges in multimodal detection by introducing three novel components: (1) the SAF-Conv module, which extends receptive field through improved submanifold sparse convolution, enhancing feature extraction from while effectively reducing edge noise; (2) Cross-modal...
Compressed video super-resolution (VSR) is employed to generate high-resolution (HR) videos from low-resolution (LR) compressed videos. Recently, some VSR methods have adopted coding priors, such as partition maps, residual frames, predictive pictures and motion vectors, HR However, these disregard the design of modules according specific characteristics information, which limits application efficiency priors. In this paper, we propose a deep network that effectively introduces priors...
Blind image super-resolution (BISR) aims to construct high-resolution from low-resolution (LR) that contains unknown degradation. Although the previous methods demonstrated impressive performance by introducing degradation representation in BISR task, there still exist two problems most of them. First, they ignore characteristics different regions when generating representation. Second, lack effective supervision on generation both and result. To solve these problems, we propose dual circle...
In this paper, we propose a temporal group alignment and fusion network to enhance the quality of compressed videos by using long-short term correlations between frames. The proposed model consists intra-group feature (IntraGFA) module, inter-group (InterGFF) enhancement (FE) module. We form pictures (GoP) selecting frames from video according their distances target enhanced frame. With grouping, composed GoP can contain either long- or short-term correlated information neighboring design...