FreeFlow: bias-free hierarchical transformer hits SOTA on optical flow
Vladislav Bargatin · hf · 2026-09-11
FreeFlow is a hierarchical transformer for optical flow estimation that eliminates task-specific inductive biases. Using window, shifted-window, and global attention, it achieves state-of-the-art accuracy, showing geometric tasks can be solved without handcrafted priors.
More from Research
- Talk at Geometry of ML 2026 shows AI finding and recommending resolutions to open math conjectures — wellecks · 2026-09-11
- Fast ViT shows strong ImageNet results; scaling runs needed next — ducha_aiki · 2026-09-11
- Loss Functions Are Scientific Assumptions: MSE Implies Gaussian Noise, Cross-Entropy Implies Bernoulli — bravo_abad · 2026-09-11
- SymKit MCP: 44 tools for AI agents to verify symbolic derivations — Foreign-Specific-604 · 2026-09-11
- Researchers: LLMs under pressure invent new languages unreadable to humans — mikeflache · 2026-09-11
- Mi-Ripple fixes ripple artifacts left by iterative AI image editing — Miyang-AI · 2026-09-11