Research map / Probabilistic and scientific modeling / Molecules and proteins
"SEA-RAFT: Simple, Efficient, Accurate RAFT for Optical Flow"
Yihan Wang, Lahav O Lipson, Jia Deng
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 |
|---|---|---|---|
| replace L1 loss with mixture-of-Laplace loss (beta1=0) (ablation) | SEA-RAFT(S) · Spring subval | EPE: 0.206 -> 0.187 | Improved |
| use naive Laplace regression instead of L1 (ablation) | SEA-RAFT(S) · Spring subval | EPE: 0.206 -> 0.217 | Hurt |
| do not fix beta1=0 in mixture of Laplace (ablation) | SEA-RAFT(S) · Spring subval | EPE: 0.187 -> 0.248 | Hurt |
| use mixture of Gaussian instead of mixture of Laplace (ablation) | SEA-RAFT(S) · Spring subval | EPE: 0.187 -> 0.210 | Hurt |
| add direct regression of initial flow (ablation) | SEA-RAFT(S) · Spring subval | EPE: 0.201 -> 0.187 | Improved |
| add TartanAir rigid-flow pre-training (ablation) | SEA-RAFT(S) · Spring subval | EPE: 0.187 -> 0.179 | Improved |
| add TartanAir pre-training (ablation) | SEA-RAFT(M) · Spring train 1px | 1px outlier rate: 4.060 -> 3.888 | Improved |
| use ImageNet pre-trained backbones (ablation) | SEA-RAFT(S) · Spring subval | EPE: 0.194 -> 0.187 | Improved |
| replace RAFT GRU with two ConvNeXt blocks (ablation) | SEA-RAFT(S) · Spring subval | EPE and MACs: EPE 0.189 -> 0.187; 297.9G -> 284.7G MACs | Improved |
| use four ConvNeXt blocks instead of two (ablation) | SEA-RAFT(S) · Spring subval | EPE: 0.187 -> 0.189 | No clear effect |
| add TartanAir pre-training (ablation) | SEA-RAFT(S) · Sintel train final (zero-shot) | EPE: 4.32 -> 3.74 | Improved |
| add small amount of KITTI and HD1K data (ablation) | SEA-RAFT(S) · Sintel train final | EPE: 4.32 -> 2.95 | Improved |
Closest papers by meaning
- MS-RAFT+: High Resolution Multi-Scale RAFT (IJCV 2023 · similarity 0.89)
- RAFT: Recurrent All-Pairs Field Transforms for Optical Flow (ECCV 2020 · similarity 0.89)
- Disentangling Architecture and Training for Optical Flow (ECCV 2022 · similarity 0.88)
- RAFT-MSF: Self-Supervised Monocular Scene Flow Using Recurrent Optimizer (IJCV 2023 · similarity 0.87)
- ZeroFlow: Scalable Scene Flow via Distillation (ICLR 2024 · similarity 0.83)
- ADFactory: An Effective Framework for Generalizing Optical Flow with NeRF (CVPR 2024 · similarity 0.83)
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.
