PaperActive

FreeFlow: A Bias-free Hierarchical Transformer for Optical Flow Estimation

quality82

Updated 4 h ago · first seen 11 Sept 2026

paper_01M294H24VPKWYWZY3AD088WJ1

Published: 11 Sept 2026; T1 · 4 h ago
arXiv: 2609.11486; T1 · 4 h ago
Category: cs.CV; T1 · 4 h ago

Abstract

Optical flow methods typically rely on task-specific inductive biases, such as correlation volumes, feature warping, and iterative refinement, among others, to reach high accuracy. While effective, such biases constrain the model to predefined heuristics, which can limit its expressivity and lead to more complex pipelines and additional computational cost. We present FreeFlow, a hierarchical transformer built without any flow-specific components, using instead a single feed-forward encoder--decoder. FreeFlow combines three attention variants: window attention for local processing, shifted-window attention for cross-window information exchange, and a global attention operating at a reduced resolution. The resulting architecture scales naturally with model capacity, enabling a consistent accuracy gain from small to large variants. Despite the absence of standard inductive biases, FreeFlow achieves state-of-the-art results on major benchmarks, including Sintel (0.68/1.48 EPE on Clean/Final), KITTI-2015 (3.23 Fl-all), and Spring (3.192 1px), while remaining memory efficient at 1080p inference.

Authors 4

Vladislav Bargatin, Alexander Yakovenko, Khaled Abud, Dmitriy Vatolin

Specification

Official page: arxiv.org/abs/2609.11486
Source:arXiv (Atom API + RSS)T1observed 4 h agohigh
Arxiv announce type: new
Source:arXiv (Atom API + RSS)T1observed 4 h agohigh
arXiv id: 2609.11486
Source:arXiv (Atom API + RSS)T1observed 4 h agohigh
Categories: cs.CV
Source:arXiv (Atom API + RSS)T1observed 4 h agohigh
DOI: 10.1007/978-3-032-37132-4_11
Source:arXiv (Atom API + RSS)T1observed 4 h agohigh
Github repo: msu-video-group/freeflow
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 h agomedium
Hf paper url: https://huggingface.co/papers/2609.11486
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 h agomedium
Github stars: 2
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 h agomedium
Hf comments: 1
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 h agomedium
Upvotes: 5
Source:Hugging Face Hub (public pages, model cards, papers)T2observed 4 h agomedium
PDF: arxiv.org/pdf/2609.11486
Source:arXiv (Atom API + RSS)T1observed 4 h agohigh
Primary category: cs.CV
Source:arXiv (Atom API + RSS)T1observed 4 h agohigh
Published: 11 Sept 2026
Source:arXiv (Atom API + RSS)T1observed 4 h agohigh

Each value shows its source, tier and observation time. Conflicting claims are kept side by side and flagged — never averaged. How AI Atlas records facts →

Provenance

Attributed facts

Source tiers

T1T210 / 5

Freshest observation

4 h ago

Conflicts

2 flagged

As of

Rewind the record: see this entity's attributes exactly as AI Atlas knew them on a given day.

Claim history

17 claims · 15 properties2 conflicting

Official pageofficial_url1

Claim history for Official page
Value	Valid from → to	Status	Source	Confidence	Extractor
https://arxiv.org/abs/2609.11486	11 Sept 2026 21:03 UTC → current	current	arXiv (Atom API + RSS)T1	high	deterministic

Abstractabstract1

Claim history for Abstract
Value	Valid from → to	Status	Source	Confidence	Extractor
Optical flow methods typically rely on task-specific inductive biases, such as correlation volumes, feature warping, and iterative refinement, among others, to reach high accuracy. While effective, such biases constrain the model to predefined heuristics, which can limit its expressivity and lead to more complex pipelines and additional computational cost. We present FreeFlow, a hierarchical transformer built without any flow-specific components, using instead a single feed-forward encoder--decoder. FreeFlow combines three attention variants: window attention for local processing, shifted-window attention for cross-window information exchange, and a global attention operating at a reduced resolution. The resulting architecture scales naturally with model capacity, enabling a consistent accuracy gain from small to large variants. Despite the absence of standard inductive biases, FreeFlow achieves state-of-the-art results on major benchmarks, including Sintel (0.68/1.48 EPE on Clean/Final), KITTI-2015 (3.23 Fl-all), and Spring (3.192 1px), while remaining memory efficient at 1080p inference.	11 Sept 2026 21:03 UTC → current	current	arXiv (Atom API + RSS)T1	high	deterministic

Arxiv announce typearxiv_announce_type1

Claim history for Arxiv announce type
Value	Valid from → to	Status	Source	Confidence	Extractor
new	11 Sept 2026 21:03 UTC → current	current	arXiv (Atom API + RSS)T1	high	deterministic

arXiv idarxiv_id1

Claim history for arXiv id
Value	Valid from → to	Status	Source	Confidence	Extractor
2609.11486	11 Sept 2026 21:03 UTC → current	current	arXiv (Atom API + RSS)T1	high	deterministic

Authorsauthors1

Claim history for Authors
Value	Valid from → to	Status	Source	Confidence	Extractor
Vladislav Bargatin, Alexander Yakovenko, Khaled Abud, Dmitriy Vatolin	11 Sept 2026 21:03 UTC → current	current	arXiv (Atom API + RSS)T1	high	deterministic

Categoriescategories1

Claim history for Categories
Value	Valid from → to	Status	Source	Confidence	Extractor
cs.CV	11 Sept 2026 21:03 UTC → current	current	arXiv (Atom API + RSS)T1	high	deterministic

DOIdoi1

Claim history for DOI
Value	Valid from → to	Status	Source	Confidence	Extractor
10.1007/978-3-032-37132-4_11	11 Sept 2026 21:03 UTC → current	current	arXiv (Atom API + RSS)T1	high	deterministic

Github repogithub_repo1

Claim history for Github repo
Value	Valid from → to	Status	Source	Confidence	Extractor
msu-video-group/freeflow	11 Sept 2026 21:10 UTC → current	current	Hugging Face Hub (public pages, model cards, papers)T2	medium	deterministic

Hf paper urlhf_paper_url1

Claim history for Hf paper url
Value	Valid from → to	Status	Source	Confidence	Extractor
https://huggingface.co/papers/2609.11486	11 Sept 2026 21:10 UTC → current	current	Hugging Face Hub (public pages, model cards, papers)T2	medium	deterministic

Github starsmetric.github_stars1

Claim history for Github stars
Value	Valid from → to	Status	Source	Confidence	Extractor
2	11 Sept 2026 21:10 UTC → current	current	Hugging Face Hub (public pages, model cards, papers)T2	medium	deterministic

Hf commentsmetric.hf_comments1

Claim history for Hf comments
Value	Valid from → to	Status	Source	Confidence	Extractor
1	11 Sept 2026 21:10 UTC → current	current	Hugging Face Hub (public pages, model cards, papers)T2	medium	deterministic

Upvotesmetric.upvotes1

Claim history for Upvotes
Value	Valid from → to	Status	Source	Confidence	Extractor
5	11 Sept 2026 21:10 UTC → current	current	Hugging Face Hub (public pages, model cards, papers)T2	medium	deterministic

PDFpdf_url1

Claim history for PDF
Value	Valid from → to	Status	Source	Confidence	Extractor
https://arxiv.org/pdf/2609.11486	11 Sept 2026 21:03 UTC → current	current	arXiv (Atom API + RSS)T1	high	deterministic

Primary categoryprimary_category1

Claim history for Primary category
Value	Valid from → to	Status	Source	Confidence	Extractor
cs.CV	11 Sept 2026 21:03 UTC → current	current	arXiv (Atom API + RSS)T1	high	deterministic

Publishedpublished_at3conflicting claims

Claim history for Published
Value	Valid from → to	Status	Source	Confidence	Extractor
10 Sept 2026	11 Sept 2026 21:12 UTC → current	conflicting	Hugging Face Hub (public pages, model cards, papers)T2	medium	deterministic
10 Sept 2026	11 Sept 2026 21:10 UTC → current	conflicting	Hugging Face Hub (public pages, model cards, papers)T2	medium	deterministic
11 Sept 2026	11 Sept 2026 21:03 UTC → current	current	arXiv (Atom API + RSS)T1	conflicted	deterministic

Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →

Source documents
Source	Document	Type	Tier	Last observed	Snapshots
arXiv (Atom API + RSS)	rss.arxiv.org/rss/cs.CV	feed	T1· Official	4 h ago	1
Hugging Face Hub (public pages, model cards, papers)	huggingface.co/papers	listing	T2· Quality secondary	4 h ago	2

Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.