pith. sign in
Pith Number

pith:DE5HB5NE

pith:2025:DE5HB5NEEANUS4WY2X3QLHYLFJ
not attested not anchored not stored refs resolved

Seg-Zero: Reasoning-Chain Guided Segmentation via Cognitive Reinforcement

Bei Yu, Bohao Peng, Fanbin Lu, Jiaya Jia, Yuqi Liu, Zhisheng Zhong, Zihao Yue

Reinforcement learning with format and accuracy rewards enables explicit reasoning chains to guide image segmentation.

arxiv:2503.06520 v2 · 2025-03-09 · cs.CV · cs.MM

Add to your LaTeX paper
\usepackage{pith}
\pithnumber{DE5HB5NEEANUS4WY2X3QLHYLFJ}

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

Seg-Zero-7B achieves a zero-shot performance of 57.5 on the ReasonSeg benchmark, surpassing the prior LISA-7B by 18%.

C2weakest assumption

That the format-plus-accuracy reward mechanism, applied only through reinforcement learning without any explicit reasoning supervision, reliably produces useful and generalizable chain-of-thought reasoning rather than superficial patterns that happen to score well on the training distribution.

C3one line summary

Seg-Zero uses cognitive reinforcement learning on a decoupled reasoning-plus-segmentation architecture to produce explicit reasoning chains and reach 57.5 zero-shot accuracy on ReasonSeg, beating prior supervised LISA-7B by 18%.

References

45 extracted · 45 resolved · 12 Pith anchors

[1] Segnet: A deep convolutional encoder-decoder architecture for image segmentation 2017
[2] Qwen2.5-VL Technical Report 2025 · arXiv:2502.13923
[3] One token to seg them all: Language instructed reasoning seg- mentation in videos 2025
[4] Deeplab: Semantic image segmentation with deep convolutional nets, atrous convolu- tion, and fully connected crfs 2017
[5] Rethinking Atrous Convolution for Semantic Image Segmentation · arXiv:1706.05587

Cited by

41 papers in Pith

Receipt and verification
First computed 2026-05-17T23:38:47.874901Z
Builder pith-number-builder-2026-05-17-v1
Signature Pith Ed25519 (pith-v1-2026-05) · public key
Schema pith-number/v1.0

Canonical hash

193a70f5a4201b4972d8d5f7059f0b2a766b9922fa756479a597657115b20c1b

Aliases

arxiv: 2503.06520 · arxiv_version: 2503.06520v2 · doi: 10.48550/arxiv.2503.06520 · pith_short_12: DE5HB5NEEANU · pith_short_16: DE5HB5NEEANUS4WY · pith_short_8: DE5HB5NE
Agent API
Verify this Pith Number yourself
curl -sH 'Accept: application/ld+json' https://pith.science/pith/DE5HB5NEEANUS4WY2X3QLHYLFJ \
  | 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: 193a70f5a4201b4972d8d5f7059f0b2a766b9922fa756479a597657115b20c1b
Canonical record JSON
{
  "metadata": {
    "abstract_canon_sha256": "75113e3d41151dbc49215b079d83c571ce82e218aa1668b01c8d75b710268a29",
    "cross_cats_sorted": [
      "cs.MM"
    ],
    "license": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
    "primary_cat": "cs.CV",
    "submitted_at": "2025-03-09T08:48:51Z",
    "title_canon_sha256": "59dc4ae3867082357577c8d924ebc5cfbdaa49794b05dbc0ca9b5cb5270779ec"
  },
  "schema_version": "1.0",
  "source": {
    "id": "2503.06520",
    "kind": "arxiv",
    "version": 2
  }
}