{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/visionreasoner-unified-visual-perception-and","title":"VisionReasoner: Unified Visual Perception and Reasoning via Reinforcement Learning","arxiv_id":"2505.12081","date":"2025-05-17","proceeding":null,"authors":["Yuqi Liu","Tianyuan Qu","Zhisheng Zhong","Bohao Peng","Shu Liu","Bei Yu","Jiaya Jia"],"abstract":"Large vision-language models exhibit inherent capabilities to handle diverse visual perception tasks. In this paper, we introduce VisionReasoner, a unified framework capable of reasoning and solving multiple visual perception tasks within a shared model. Specifically, by designing novel multi-object cognitive learning strategies and systematic task reformulation, VisionReasoner enhances its reasoning capabilities to analyze visual inputs, and addresses diverse perception tasks in a unified framework. The model generates a structured reasoning process before delivering the desired outputs responding to user queries. To rigorously assess unified visual perception capabilities, we evaluate VisionReasoner on ten diverse tasks spanning three critical domains: detection, segmentation, and counting. Experimental results show that VisionReasoner achieves superior performance as a unified model, outperforming Qwen2.5VL by relative margins of 29.1% on COCO (detection), 22.1% on ReasonSeg (segmentation), and 15.3% on CountBench (counting).","url_abs":"https://arxiv.org/abs/2505.12081v2","url_pdf":"https://arxiv.org/pdf/2505.12081v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"visionreasoner-unified-visual-perception-and","repo_url":"https://github.com/dvlab-research/VisionReasoner","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"visionreasoner-unified-visual-perception-and","repo_url":"https://github.com/hiyouga/easyr1","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"visionreasoner-unified-visual-perception-and","repo_url":"https://github.com/dvlab-research/Seg-Zero","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[{"task_slug":"2d-object-detection","task_name":"2D Object Detection"},{"task_slug":"object-counting","task_name":"Object Counting"},{"task_slug":"reasoning-segmentation","task_name":"Reasoning Segmentation"},{"task_slug":"referring-expression-segmentation","task_name":"Referring Expression Segmentation"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"zero-shot-object-detection","task_name":"Zero-Shot Object Detection"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2505.12081","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2505.12081"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/dvlab-research/VisionReasoner","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/hiyouga/easyr1","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/dvlab-research/Seg-Zero","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran_draft_wrong":1,"unverified":14},"by_repo_kind":{"official":{"samples":15,"ran":1,"repositories":2}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"32cbd7aa87e2f64b","entry":"merge_by_placement","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"training_scripts/model_merger.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/training_scripts/model_merger.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"32cbd7aa87e2f64b"}},{"code_sha256_prefix":"cea027a0a2a1de9b","entry":"compute_bbox_iou","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"evaluation_scripts/evaluation_visionreasoner.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/evaluation_scripts/evaluation_visionreasoner.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"cea027a0a2a1de9b"}},{"code_sha256_prefix":"43902f82943491dd","entry":"compute_bbox_iou","repo":"dvlab-research/VisionReasoner","repo_kind":"official","path":"evaluation/evaluation_coco.py","file_url":"https://github.com/dvlab-research/VisionReasoner/blob/HEAD/evaluation/evaluation_coco.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"43902f82943491dd"}},{"code_sha256_prefix":"faee26847645d13b","entry":"compute_iou","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"evaluation_scripts/evaluation.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/evaluation_scripts/evaluation.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"faee26847645d13b"}},{"code_sha256_prefix":"611c245861107d30","entry":"compute_iou","repo":"dvlab-research/VisionReasoner","repo_kind":"official","path":"evaluation/evaluation_anomaly.py","file_url":"https://github.com/dvlab-research/VisionReasoner/blob/HEAD/evaluation/evaluation_anomaly.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"611c245861107d30"}},{"code_sha256_prefix":"9b81b23f39e8998d","entry":"create_and_push_dataset","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"evaluation_scripts/create_reasonseg_hf.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/evaluation_scripts/create_reasonseg_hf.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"9b81b23f39e8998d"}},{"code_sha256_prefix":"ed9a13dcc8178311","entry":"extract_bbox_points_think","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"evaluation_scripts/evaluation.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/evaluation_scripts/evaluation.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"ed9a13dcc8178311"}},{"code_sha256_prefix":"732abc1b7bf1a77b","entry":"extract_bbox_points_think","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"evaluation_scripts/evaluation_visionreasoner.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/evaluation_scripts/evaluation_visionreasoner.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"732abc1b7bf1a77b"}},{"code_sha256_prefix":"e17c29f2dfb12d4a","entry":"extract_bbox_points_think","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"inference_scripts/infer_multi_object.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/inference_scripts/infer_multi_object.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e17c29f2dfb12d4a"}},{"code_sha256_prefix":"af0760e7da48075b","entry":"get_mask_from_json","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"evaluation_scripts/create_reasonseg_hf.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/evaluation_scripts/create_reasonseg_hf.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"af0760e7da48075b"}},{"code_sha256_prefix":"58dcb3747ca297c1","entry":"seg_iou_reward","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"verl/utils/reward_score/seg.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/verl/utils/reward_score/seg.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"58dcb3747ca297c1"}},{"code_sha256_prefix":"80bd3e5670352790","entry":"seg_iou_reward","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"verl/utils/reward_score/seg_restrict.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/verl/utils/reward_score/seg_restrict.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"80bd3e5670352790"}},{"code_sha256_prefix":"77f785f63d30a16b","entry":"seg_segmentation_format_reward","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"verl/utils/reward_score/seg.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/verl/utils/reward_score/seg.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"77f785f63d30a16b"}},{"code_sha256_prefix":"0d43353f3a3abfff","entry":"seg_segmentation_format_reward","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"verl/utils/reward_score/seg_restrict.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/verl/utils/reward_score/seg_restrict.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"0d43353f3a3abfff"}},{"code_sha256_prefix":"a0fec9aec44b558c","entry":"seg_thinking_format_reward","repo":"dvlab-research/Seg-Zero","repo_kind":"official","path":"verl/utils/reward_score/seg.py","file_url":"https://github.com/dvlab-research/Seg-Zero/blob/HEAD/verl/utils/reward_score/seg.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"a0fec9aec44b558c"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}