Summary of High-frequency Enhanced Hybrid Neural Representation For Video Compression, by Li Yu et al.
High-Frequency Enhanced Hybrid Neural Representation for Video Compression
by Li Yu, Zhihui Li, Jimin Xiao, Moncef Gabbouj
First submitted to arxiv on: 11 Nov 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Image and Video Processing (eess.IV)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary Neural Representations for Videos (NeRV) have simplified video codec processes by encoding content into neural networks, promising a solution for video compression. However, existing work overlooks the issue of reconstructed videos lacking high-frequency details. This paper introduces a High-Frequency Enhanced Hybrid Neural Representation Network that leverages high-frequency information to improve fine detail synthesis. The method combines wavelet high-frequency encoders with Wavelet Frequency Decomposer (WFD) blocks and High-Frequency Feature Modulation (HFM) blocks to refine the decoder. The Harmonic decoder block and Dynamic Weighted Frequency Loss further reduce potential losses. Experiments on Bunny and UVG datasets demonstrate improvements in detail preservation and compression performance. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Videos compressed with Neural Representations for Videos (NeRV) have become faster to decode, but there’s a problem: the restored videos lack important details. To fix this, scientists created a new way to compress video that keeps these details intact. They used special blocks called wavelet high-frequency encoders and Wavelet Frequency Decomposer (WFD) blocks to add more detail back into the compressed video. Then, they used another block called High-Frequency Feature Modulation (HFM) to make sure all the important parts of the video were preserved. Finally, they tested their new method on two different datasets and found that it worked much better than other methods. |
Keywords
» Artificial intelligence » Decoder
