pith. sign in
Pith Number

pith:OO2D3RD6

pith:2025:OO2D3RD6XHEATH5NY5DU322KDF
not attested not anchored not stored refs resolved

VideoChat-R1: Enhancing Spatio-Temporal Perception via Reinforcement Fine-Tuning

Desen Meng, Limin Wang, Lu Dong, Xiangyu Zeng, Xinhao Li, Yali Wang, Yinan He, Yi Wang, Yu Qiao, Ziang Yan

Reinforcement fine-tuning with rule-based temporal rewards creates a video model with state-of-the-art spatio-temporal perception.

arxiv:2504.06958 v5 · 2025-04-09 · cs.CV

Add to your LaTeX paper
\usepackage{pith}
\pithnumber{OO2D3RD6XHEATH5NY5DU322KDF}

Prints a linked badge after your title and injects PDF metadata. Compiles on arXiv. Learn more · Embed verified badge

Record completeness

1 Bitcoin timestamp
2 Internet Archive
3 Author claim open · sign in to claim
4 Citations open
5 Replications open
Portable graph bundle live · download bundle · merged state
The bundle contains the canonical record plus signed events. A mirror can host it anywhere and recompute the same current state with the deterministic merge algorithm.

Claims

C1strongest claim

Through joint RFT on multiple spatio-temporal perception tasks, we developed VideoChat-R1, a powerful Video MLLM. VideoChat-R1 achieves state-of-the-art spatio-temporal perception, demonstrating significant improvements in tasks like temporal grounding (+31.8) and object tracking (+31.2), while also improving general QA benchmarks.

C2weakest assumption

That carefully designed rule-based rewards focused on temporal associations will produce generalizable improvements in video reasoning without post-hoc tuning or hidden data selection that inflates the reported deltas.

C3one line summary

Reinforcement fine-tuning with temporal rewards produces VideoChat-R1, a video MLLM showing large gains on spatio-temporal perception benchmarks such as +31.8 temporal grounding and +31.2 object tracking.

References

42 extracted · 42 resolved · 19 Pith anchors

[1] Qwen2.5-VL Technical Report 2025 · arXiv:2502.13923
[2] et al.: FlashVTG: Feature layering and adaptive score handling network for video temporal grounding 2024
[3] Boosting the generalization and reasoning of vision language models with curriculum reinforcement learning 2025
[4] Open- vlthinker: Complex vision-language reasoning via iterative sft-rl cycles 2025
[5] Video-R1: Reinforcing Video Reasoning in MLLMs 2025 · arXiv:2503.21776

Formal links

2 machine-checked theorem links

Cited by

46 papers in Pith

Receipt and verification
First computed 2026-05-17T23:38:50.192174Z
Builder pith-number-builder-2026-05-17-v1
Signature Pith Ed25519 (pith-v1-2026-05) · public key
Schema pith-number/v1.0

Canonical hash

73b43dc47eb9c8099fadc7474deb4a195144c4997a545220f336e76bc37dbcc5

Aliases

arxiv: 2504.06958 · arxiv_version: 2504.06958v5 · doi: 10.48550/arxiv.2504.06958 · pith_short_12: OO2D3RD6XHEA · pith_short_16: OO2D3RD6XHEATH5N · pith_short_8: OO2D3RD6
Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/OO2D3RD6XHEATH5NY5DU322KDF \
  | jq -c '.canonical_record' \
  | python3 -c "import sys,json,hashlib; b=json.dumps(json.loads(sys.stdin.read()), sort_keys=True, separators=(',',':'), ensure_ascii=False).encode(); print(hashlib.sha256(b).hexdigest())"
# expect: 73b43dc47eb9c8099fadc7474deb4a195144c4997a545220f336e76bc37dbcc5
Canonical record JSON
{
  "metadata": {
    "abstract_canon_sha256": "67904bdf7f730bb66fe9d28a566e7454555ba6eba27c2c7f6e7a7cc7701082ef",
    "cross_cats_sorted": [],
    "license": "http://creativecommons.org/licenses/by/4.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2025-04-09T15:09:27Z",
    "title_canon_sha256": "ec44830ff7b27ba891b808dca66f374e132a4205f905cac9be95422336277a31"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2504.06958",
    "kind": "arxiv",
    "version": 5
  }
}