Research map / Video / Video segmentation and tracking
Tree-NeRV: Efficient Non-Uniform Sampling for Neural Video Representation via Tree-Structured Feature Grids
Jiancheng Zhao, Yifan Zhan, Qingtian Zhu, Mingze Ma, Muyao Niu, Zunian Wan, Xiang Ji, Yinqiang Zheng
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use adaptive tree-growing sampling instead of uniform sampling at 100 feature points (ablation) | Tree-NeRV · UVG | PSNR (dB): 32.11 -> 33.36 | Improved |
| replace uniform grid with tree-structured grid with non-uniform sampling | feature-grid NeRV · UVG (avg over methods) | PSNR (dB): +0.73dB over competing methods | Improved |
| replace uniform grid with tree-structured grid with non-uniform sampling | feature-grid NeRV · DAVIS | PSNR (dB): +1.26dB over competing methods | Improved |
| tree-structured grid for video interpolation | Tree-NeRV vs DS-NeRV · UVG interpolation | PSNR (dB): 29.38 -> 30.16 | Improved |
| tree-structured grid at small model size on Bunny | Tree-NeRV · Bunny 720P (0.35M) | PSNR (dB): 29.78 -> 30.21 | Improved |
| tree-structured grid reduces encoding time versus feature-grid baselines | Tree-NeRV vs DS-NeRV · UVG | encoding time (h): ~3h -> ~2h | Improved |
| tree query and interpolation at decoding versus per-frame latent (HNeRV) | Tree-NeRV vs HNeRV · UVG | decoding FPS: 105.90 -> 70.14 | Hurt |
| train longer (5 min to 2 hours) | Tree-NeRV · UVG | PSNR (dB): 28.00 -> 33.36 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- DS-NeRV — DS-NeRV: Implicit Neural Video Representation with Decomposed Static and Dynamic Codes (CVPR 2024) · better on 9 of 9 shared results
- HNeRV — HNeRV: A Hybrid Neural Representation for Videos (CVPR 2023) · better on 3 of 4 shared results
- NeRV — NeRV: Neural Representations for Videos (NeurIPS 2021) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- DS-NeRV: Implicit Neural Video Representation with Decomposed Static and Dynamic Codes (CVPR 2024 · similarity 0.90)
- Combining Frame and GOP Embeddings for Neural Video Representation (CVPR 2024 · similarity 0.89)
- Boosting Neural Representations for Videos with a Conditional Decoder (CVPR 2024 · similarity 0.88)
- PNeRV: Enhancing Spatial Consistency via Pyramidal Neural Representation for Videos (CVPR 2024 · similarity 0.88)
- MetaNeRV: Meta Neural Representations for Videos with Spatial-Temporal Guidance (AAAI 2025 · similarity 0.88)
- Neural Video Compression with Feature Modulation (CVPR 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
