{"id":31570357,"url":"https://github.com/sairajk/asia","last_synced_at":"2026-04-18T10:36:56.339Z","repository":{"id":316997685,"uuid":"1065578795","full_name":"sairajk/asia","owner":"sairajk","description":"[SIGGRAPH Asia 2025] \"ASIA: Adaptive 3D Segmentation using Few Image Annotations \".","archived":false,"fork":false,"pushed_at":"2026-02-14T02:54:44.000Z","size":73297,"stargazers_count":22,"open_issues_count":0,"forks_count":0,"subscribers_count":4,"default_branch":"main","last_synced_at":"2026-02-14T09:05:24.054Z","etag":null,"topics":["3d","controlnet","deep-learning","few-shot","few-shot-segmentation","machine-learning","mesh-segmentation","segmentation","stable-diffusion","test-time-optimization"],"latest_commit_sha":null,"homepage":"https://sairajk.github.io/asia/","language":null,"has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/sairajk.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-09-28T02:20:08.000Z","updated_at":"2025-12-09T00:31:59.000Z","dependencies_parsed_at":"2025-09-28T06:29:31.439Z","dependency_job_id":null,"html_url":"https://github.com/sairajk/asia","commit_stats":null,"previous_names":["sairajk/asia"],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/sairajk/asia","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sairajk%2Fasia","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sairajk%2Fasia/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sairajk%2Fasia/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sairajk%2Fasia/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/sairajk","download_url":"https://codeload.github.com/sairajk/asia/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/sairajk%2Fasia/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31966206,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-18T00:39:45.007Z","status":"online","status_checked_at":"2026-04-18T02:00:07.018Z","response_time":103,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["3d","controlnet","deep-learning","few-shot","few-shot-segmentation","machine-learning","mesh-segmentation","segmentation","stable-diffusion","test-time-optimization"],"created_at":"2025-10-05T12:21:30.207Z","updated_at":"2026-04-18T10:36:56.330Z","avatar_url":"https://github.com/sairajk.png","language":null,"funding_links":[],"categories":[],"sub_categories":[],"readme":"\u003cp align=\"center\"\u003e\n\n  \u003ch1 align=\"center\"\u003e\u003ca href=\"https://sairajk.github.io/asia\" target=\"_blank\"\u003eASIA: Adaptive 3D Segmentation using Few Image Annotations\u003c/a\u003e\u003c/h1\u003e\n\n  \u003cp align=\"center\"\u003e\n    \u003ca href=\"https://sairajk.github.io/\" target=\"_blank\"\u003e\u003cstrong\u003eSai Raj Kishore Perla\u003c/strong\u003e\u003c/a\u003e\n    ·\n    \u003ca href=\"https://aditya-vora.github.io/\" target=\"_blank\"\u003e\u003cstrong\u003eAditya Vora\u003c/strong\u003e\u003c/a\u003e\n    ·\n    \u003ca href=\"https://sauradip.github.io/\" target=\"_blank\"\u003e\u003cstrong\u003eSauradip Nag\u003c/strong\u003e\u003c/a\u003e\n    ·\n    \u003ca href=\"https://www.sfu.ca/~amahdavi/\" target=\"_blank\"\u003e\u003cstrong\u003eAli Mahdavi-Amiri\u003c/strong\u003e\u003c/a\u003e\n    ·\n    \u003ca href=\"https://www.cs.sfu.ca/~haoz/\" target=\"_blank\"\u003e\u003cstrong\u003eHao (Richard) Zhang\u003c/strong\u003e\u003c/a\u003e\n    \u003cbr /\u003e\n    \u003ci\u003eSIGGRAPH Asia, 2025\u003c/i\u003e\n  \u003c/p\u003e\n\n  \u003cp align=\"center\"\u003e\n    \u003ca href=\"https://arxiv.org/abs/\" target=\"_blank\"\u003e\u003cstrong\u003earXiv\u003c/strong\u003e\u003c/a\u003e\n    |\n    \u003ca href=\"https://sairajk.github.io/asia\" target=\"_blank\"\u003e\u003cstrong\u003eProject Page\u003c/strong\u003e\u003c/a\u003e\n  \u003c/p\u003e\n\n  \u003cdiv  align=\"center\"\u003e\n    \u003ca\u003e\n      \u003cimg src=\"https://github.com/sairajk/asia/blob/gh-pages/static/assets/teaser_v3.png\" alt=\"Logo\" width=\"100%\"\u003e\n    \u003c/a\u003e\n  \u003c/div\u003e\n\u003c/p\u003e\n\nWe introduce ASIA (Adaptive 3D Segmentation using few Image Annotations), a novel framework that enables segmentation of possibly non-semantic and non-text-describable “parts” in 3D. Our segmentation is controllable through a few user-annotated in-the-wild images, which are easier to collect than multi-view images, less demanding to annotate than 3D models, and more precise than potentially ambiguous text descriptions. Our method leverages the rich priors of text-to-image diffusion models, such as Stable Diffusion (SD), to transfer segmentations from image space to 3D, even when the annotated and target objects differ significantly in geometry or structure. During training, we optimize a text token for each segment and fine-tune our model with a novel cross-view part correspondence loss. At inference, we segment multi-view renderings of the 3D mesh, fuse the labels in UV-space via voting, refine them with our novel Noise Optimization technique, and finally map the UV-labels back onto the mesh. ASIA provides a practical and generalizable solution for both semantic and non-semantic 3D segmentation tasks, outperforming existing methods by a noticeable margin in both quantitative and qualitative evaluations.\n\n\u003cbr\u003e\n\u003ch3 align=\"center\"\u003e🚧 Code will be released soon. 🚧\u003c/h3\u003e\n\u003cbr\u003e\n\n## Citation\n\nIf you found our work helpful, please consider citing:\n\n```bibtex\n@article{perla2025asia,\n    title={{ASIA}: Adaptive 3D Segmentation using Few Image Annotations},\n    author = {Perla, Sai Raj Kishore and Vora, Aditya and Nag, Sauradip, Mahdavi-Amiri, Ali and Zhang, Hao},\n    journal = {SIGGRAPH Asia Conference Papers},\n    publisher = {ACM New York, NY, USA},\n    year = {2025},\n    doi = {10.1145/3757377.3763821},\n    url = {https://github.com/sairajk/asia},\n}\n```\n\n\u003c!-- volume = {43},\n    number = {4},\n    articleno = {40}, --\u003e\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsairajk%2Fasia","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fsairajk%2Fasia","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fsairajk%2Fasia/lists"}