Co-Via: A Video Frame Interpolation Accelerator Exploiting Codec Information Reuse
Haishuang Fan, Qichu Sun, Jingya Wu, Wenyan Lu, Xiaowei Li, Guihai Yan
Abstract
Video Frame Interpolation (VFI) aims to generate intermediate frames between consecutive frames. Recent DNN-based VFI offers superior quality but suffers from performance issues. However, very few studies have focused on VFI hardware acceleration and existing work overlooks temporal information from compressed video bitstreams. In this paper, we propose a novel compressed VFI workflow and an accelerator, Co-Via. Co-Via exploits codec information reuse to reduce complex DNN computations and alleviate hardware pressure. FPGA-based Co-Via outperforms an RTX 4090 GPU 10.31X, offering a 43.08X energy efficiency boost. Its ASIC version achieves 2.4X higher throughput and 3.6X energy efficiency than the state-of-the-art solution.
Ask about this paper
Ask your agent about it.
Lune has read the top-tier papers around this one, so every answer names the papers it rests on.
Your agent calls
Lunesearch_papers
Free to start. No credit card required.
Terminal
Install the CLIlune papers get 94263de6-ecea-43a5-82fd-21b4fc1bc1ffRelated papers
- Extracting Motion and Appearance via Inter-Frame Attention for Efficient Video Frame InterpolationGuozhen Zhang, Yuhan Zhu, Haonan Wang, Youxin Chen et al.CVPR 2023
- DeltaCNN: End-to-End CNN Inference of Sparse Frame Differences in VideosMathias Parger, Chengcheng Tang, Christopher D. Twigg, Cem Keskin et al.CVPR 2022 · 31 citations
- An Efficient Deep Learning Accelerator for Compressed Video AnalysisYongchen Wang, Ying Wang, Huawei Li, Yinhe Han et al.DAC 2020 · 4 citations
- Hierarchical Flow Diffusion for Efficient Frame InterpolationYang Hai, Guo Wang, Tan Su, Wenjie Jiang et al.CVPR 2025
- GoSPA: An Energy-efficient High-performance Globally Optimized SParse Convolutional Neural Network AcceleratorChunhua Deng, Yang Sui, Siyu Liao, Xuehai Qian et al.ISCA 2021 · 77 citations
