{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T15:15:59Z","timestamp":1785942959338,"version":"3.56.0"},"reference-count":33,"publisher":"Wiley","license":[{"start":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T00:00:00Z","timestamp":1785888000000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/onlinelibrary.wiley.com\/termsAndConditions#vor"},{"start":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T00:00:00Z","timestamp":1785888000000},"content-version":"tdm","delay-in-days":0,"URL":"http:\/\/doi.wiley.com\/10.1002\/tdm_license_1.1"}],"content-domain":{"domain":["onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Computer Graphics Forum"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>Creating traditional 3D scenes is both time\u2010consuming and costly, requiring designers to meticulously configure 3D assets and environments. Recent advancements in generative AI, including text\u2010to\u20103D and image\u2010to\u20103D methods, have significantly reduced the complexity and cost of this process. However, current techniques for editing complex 3D scenes still rely heavily on interactive, multi\u2010step 2D\u2010to\u20103D projection methods and diffusion techniques, which often lack precision. Therefore, this study proposes AgentEditor \u2014\u2014 a controllable 3D scene editing and multimodal interaction framework that integrates 3D Gaussian splatting with large language models (LLMs). It establishes an LLM\u2010driven \u201cedit\u2010evaluate\u2010optimize\u201d autonomous closed\u2010loop feedback mechanism capable of quantitatively assessing semantic degradation after editing and dynamically triggering adaptive local semantic fine\u2010tuning. Extensive experimental results demonstrate that AgentEditor achieves superior editing accuracy and speed compared to current state\u2010of\u2010the\u2010art 3D scene editing methods, setting a new benchmark for efficient interactive 3D scene customization.<\/jats:p>","DOI":"10.1111\/cgf.70532","type":"journal-article","created":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T13:10:11Z","timestamp":1785935411000},"update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Instructable 3D Scene Editing via LLM\u2010Driven Closed\u2010Loop Gaussian Splatting"],"prefix":"10.1111","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-9027-3261","authenticated-orcid":false,"given":"H.","family":"Zhang","sequence":"first","affiliation":[{"name":"School of Artificial Intelligence and Data Science Shanghai Dianji University  Shanghai 201306 China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0002-0459-4785","authenticated-orcid":false,"given":"Y. W.","family":"Wu","sequence":"additional","affiliation":[{"name":"School of Artificial Intelligence and Data Science Shanghai Dianji University  Shanghai 201306 China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2691-034X","authenticated-orcid":false,"given":"Y. L.","family":"Liu","sequence":"additional","affiliation":[{"name":"School of Artificial Intelligence and Data Science Shanghai Dianji University  Shanghai 201306 China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2026,8,5]]},"reference":[{"key":"e_1_2_7_2_2","doi-asserted-by":"crossref","unstructured":"BrooksT. HolynskiA. EfrosA. A.: Instructpix2pix: Learning to follow image editing instructions. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2023). In press. 1","DOI":"10.1109\/CVPR52729.2023.01764"},{"key":"e_1_2_7_3_2","doi-asserted-by":"crossref","unstructured":"BaoC. ZhangY. YangB. et al.: Sine: Semantic\u2010driven image\u2010based nerf editing with prior\u2010guided editing field. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2023) pp.20919\u201320929. 1 2","DOI":"10.1109\/CVPR52729.2023.02004"},{"key":"e_1_2_7_4_2","doi-asserted-by":"crossref","unstructured":"ChenY. ChenZ. ZhangC. et al.: Gaussianeditor: Swift and controllable 3d editing with gaussian splatting. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2024) pp.21476\u201321485. 1 2 8 9 10","DOI":"10.1109\/CVPR52733.2024.02029"},{"key":"e_1_2_7_5_2","doi-asserted-by":"crossref","unstructured":"CenJ. FangJ. YangC. et al.: Segment any 3d gaussians. InProc. AAAI Conf. Artif. Intell. (2025) pp.1971\u20131979. 2","DOI":"10.1609\/aaai.v39i2.32193"},{"key":"e_1_2_7_6_2","doi-asserted-by":"crossref","unstructured":"ChenY. PanY. LiY. et al.: Control3d: Towards controllable text\u2010to\u20103d generation. InProc. 31st ACM Int. Conf. Multimedia(2023) pp.1148\u20131156. 1","DOI":"10.1145\/3581783.3612489"},{"key":"e_1_2_7_7_2","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2023.3305295"},{"key":"e_1_2_7_8_2","doi-asserted-by":"crossref","unstructured":"ChenZ. WangF. WangY. LiuH.: Text\u2010to\u20103d using gaussian splatting. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2024) pp.21401\u201321412. 1","DOI":"10.1109\/CVPR52733.2024.02022"},{"key":"e_1_2_7_9_2","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2686"},{"key":"e_1_2_7_10_2","doi-asserted-by":"crossref","unstructured":"Fridovich\u2010KeilS. YuA. TancikM. et al.: Plenoxels: Radiance fields without neural networks. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2022). In press. 2","DOI":"10.1109\/CVPR52688.2022.00542"},{"key":"e_1_2_7_11_2","doi-asserted-by":"crossref","unstructured":"GaoW. AigermanN. GroueixT. et al.: Textde\u2010former: Geometry manipulation using text guidance. InProc. ACM SIGGRAPH Conf. (2023). In press. 2","DOI":"10.1145\/3588432.3591552"},{"key":"e_1_2_7_12_2","doi-asserted-by":"crossref","unstructured":"HaqueA. TancikM. EfrosA. A. et al.: Instructnerf2nerf: Editing 3d scenes with instructions. InProc. IEEE\/CVF Int. Conf. Comput. Vis. (2023). In press. 1 8 9","DOI":"10.1109\/ICCV51070.2023.01808"},{"key":"e_1_2_7_13_2","doi-asserted-by":"publisher","DOI":"10.52202\/075280-0900"},{"key":"e_1_2_7_14_2","doi-asserted-by":"crossref","unstructured":"HuangT. ZengY. ZhangZ. et al.: Dreamcontrol: Control\u2010based text\u2010to\u20103d generation with 3d self\u2010prior. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2024) pp.5364\u20135373. 1","DOI":"10.1109\/CVPR52733.2024.00513"},{"key":"e_1_2_7_15_2","doi-asserted-by":"publisher","DOI":"10.1145\/3585499"},{"key":"e_1_2_7_16_2","unstructured":"KerrJ. KimC. M. GoldbergK. et al.: Lerf: Language embedded radiance fields. InProc. IEEE\/CVF Int. Conf. Comput. Vis. (2023) pp.19729\u201319739. 8"},{"key":"e_1_2_7_17_2","doi-asserted-by":"publisher","DOI":"10.1145\/3592433"},{"key":"e_1_2_7_18_2","doi-asserted-by":"crossref","unstructured":"LiuF. WuD. WeiY. et al.: Sherpa3d: Boosting high\u2010fidelity text\u2010to\u20103d generation via coarse 3d prior. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2024) pp.20763\u201320774. 1","DOI":"10.1109\/CVPR52733.2024.01962"},{"key":"e_1_2_7_19_2","doi-asserted-by":"publisher","DOI":"10.52202\/075280-2325"},{"key":"e_1_2_7_20_2","doi-asserted-by":"publisher","DOI":"10.1145\/3528223.3530127"},{"key":"e_1_2_7_21_2","doi-asserted-by":"publisher","DOI":"10.1145\/3503250"},{"key":"e_1_2_7_22_2","doi-asserted-by":"publisher","DOI":"10.52202\/079017-0810"},{"key":"e_1_2_7_23_2","unstructured":"NoguchiA. SunX. LinS. HaradaT.: Neural articulated radiance field. InProc. IEEE\/CVF Int. Conf. Comput. Vis. (2021) pp.5762\u20135772. 2"},{"key":"e_1_2_7_24_2","doi-asserted-by":"crossref","unstructured":"QinM. LiW. ZhouJ. et al.: Langsplat: 3d language gaussian splatting. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2024). In press. 2","DOI":"10.1109\/CVPR52733.2024.01895"},{"key":"e_1_2_7_25_2","doi-asserted-by":"crossref","unstructured":"RombachR. BlattmannA. LorenzD. et al.: High\u2010resolution image synthesis with latent diffusion models. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2022) pp.10674\u201310685. 1","DOI":"10.1109\/CVPR52688.2022.01042"},{"key":"e_1_2_7_26_2","doi-asserted-by":"crossref","unstructured":"ShiJ.\u2010C. WangM. DuanH.\u2010B. GuanS.\u2010H.: Language embedded 3d gaussians for open\u2010vocabulary scene understanding. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2024). In press. 2","DOI":"10.1109\/CVPR52733.2024.00510"},{"key":"e_1_2_7_27_2","first-page":"15","volume":"6","author":"Vachha C.","year":"2024","journal-title":"Instruct\u2010gs2gs: Editing 3d gaussian splats with instructions"},{"key":"e_1_2_7_28_2","doi-asserted-by":"crossref","unstructured":"WangJ. FangJ. ZhangX. et al.: Gaussianeditor: Editing 3d gaussians delicately with text instructions. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2024) pp.20902\u201320911. 1","DOI":"10.1109\/CVPR52733.2024.01975"},{"key":"e_1_2_7_29_2","unstructured":"WangZ. HuangH. ZhaoY. et al.: Chat\u20103d: Data\u2010efficiently tuning large language model for universal dialogue of 3d scenes.arXiv preprint arXiv:230808769(2023). 2"},{"key":"e_1_2_7_30_2","doi-asserted-by":"publisher","DOI":"10.1109\/TVCG.2023.3283400"},{"key":"e_1_2_7_31_2","doi-asserted-by":"publisher","DOI":"10.1021\/acs.nanolett.0c02107"},{"key":"e_1_2_7_32_2","first-page":"597","volume-title":"Proc. Eur. Conf. Comput. Vis","author":"Yang B.","year":"2022"},{"key":"e_1_2_7_33_2","first-page":"10610443","volume-title":"Proc. IEEE Int. Conf. Robot. Autom. (ICRA)","author":"Yang J.","year":"2024"},{"key":"e_1_2_7_34_2","unstructured":"YuanY.\u2010J. SunY.\u2010T. LaiY.\u2010K. et al.: Nerf\u2010editing: geometry editing of neural radiance fields. InProc. IEEE\/CVF Conf. Comput. Vis. Pattern Recognit. (2022) pp.18353\u201318364. 2"}],"container-title":["Computer Graphics Forum"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1111\/cgf.70532","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/full-xml\/10.1111\/cgf.70532","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/pdf\/10.1111\/cgf.70532","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,5]],"date-time":"2026-08-05T14:37:12Z","timestamp":1785940632000},"score":1,"resource":{"primary":{"URL":"https:\/\/onlinelibrary.wiley.com\/doi\/10.1111\/cgf.70532"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,8,5]]},"references-count":33,"alternative-id":["10.1111\/cgf.70532"],"URL":"https:\/\/doi.org\/10.1111\/cgf.70532","archive":["Portico"],"relation":{},"ISSN":["0167-7055","1467-8659"],"issn-type":[{"value":"0167-7055","type":"print"},{"value":"1467-8659","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,5]]},"assertion":[{"value":"2026-08-05","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"e70532"}}