Dr. NAWAZ Mehmood2026-06-182026-06-182025Nawaz, M. (26 Aug 2025). Deep neural networks for block-level video enhancement: Transforming LR AVS to HR HEVC video. AI-SI 2025 - IEEE International Conference on Artificial Intelligence for Sustainable Innovation: Shaping the Future with Intelligent Solutions, Seri Pacific Hotel, Kuala Lumpur, Malaysia.http://hdl.handle.net/20.500.11861/27543Converting one video bitstream to another video bitstream is a challenging task in the heterogeneous transcoder due to different video formats. In this paper, we propose a region of interest (ROI)-based super-resolution technique to convert low-resolution AVS video to high-definition HEVC video. The proposed method uses visual characteristics, transform coefficients,and motion vectors to classify a low-resolution video frame into small blocks, which are further classified as blocks of most interest (BOMI), blocks of less interest (BOLI), and blocks of noninterest (BONI). The BONI blocks are considered background blocks and remain unchanged. A deep learning-based super resolution (SR) method is then applied to the BOMI and BOLI blocks to enhance the visual quality. The proposed method saves 20 - 30% computational time and obtains appreciable results compared to the original low-resolution frames. We have tested our method on different official video sequences with resolutions of 1K, 2K, and 4K. Our method has efficient visual performance compared to the other methods.enConference Paper