000147897 001__ 147897 000147897 005__ 20250109162738.0 000147897 0247_ $$2doi$$a10.1145/3641519.3657444 000147897 0248_ $$2sideral$$a141304 000147897 037__ $$aART-2024-141304 000147897 041__ $$aeng 000147897 100__ $$0(orcid)0000-0002-2077-683X$$aGuerrero-Viu, Julia$$uUniversidad de Zaragoza 000147897 245__ $$aTexSliders: Diffusion-Based Texture Editing in CLIP Space 000147897 260__ $$c2024 000147897 5203_ $$aGenerative models have enabled intuitive image creation and manipulation using natural language. In particular, diffusion models have recently shown remarkable results for natural image editing. In this work, we propose to apply diffusion techniques to edit textures, a specific class of images that are an essential part of 3D content creation pipelines. We analyze existing editing methods and show that they are not directly applicable to textures, since their common underlying approach, manipulating attention maps, is unsuitable for the texture domain. To address this, we propose a novel approach that instead manipulates CLIP image embeddings to condition the diffusion generation. We define editing directions using simple text prompts (e.g., “aged wood” to “new wood”) and map these to CLIP image embedding space using a texture prior, with a sampling-based approach that gives us identity-preserving directions in CLIP space. To further improve identity preservation, we project these directions to a CLIP subspace that minimizes identity variations resulting from entangled texture attributes. Our editing pipeline facilitates the creation of arbitrary sliders using natural language prompts only, with no ground-truth annotated data necessary. 000147897 536__ $$9info:eu-repo/grantAgreement/EC/H2020/956585/EU/Predictive Rendering In Manufacture and Engineering/PRIME$$9This project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No H2020 956585-PRIME$$9info:eu-repo/grantAgreement/ES/MCIU/FPU20-02340$$9info:eu-repo/grantAgreement/ES/MICINN/PID2022-141766OB-I00 000147897 540__ $$9info:eu-repo/semantics/closedAccess$$aAll rights reserved$$uhttp://www.europeana.eu/rights/rr-f/ 000147897 655_4 $$ainfo:eu-repo/semantics/article$$vinfo:eu-repo/semantics/publishedVersion 000147897 700__ $$aHasan, Milos 000147897 700__ $$aRoullier, Arthur 000147897 700__ $$aHarikumar, Midhun 000147897 700__ $$aHu, Yiwei 000147897 700__ $$aGuerrero, Paul 000147897 700__ $$0(orcid)0000-0002-7503-7022$$aGutiérrez, Diego$$uUniversidad de Zaragoza 000147897 700__ $$0(orcid)0000-0003-0060-7278$$aMasia, Belen$$uUniversidad de Zaragoza 000147897 700__ $$aDeschaintre, Valentin 000147897 7102_ $$15007$$2570$$aUniversidad de Zaragoza$$bDpto. Informát.Ingenie.Sistms.$$cÁrea Lenguajes y Sistemas Inf. 000147897 773__ $$g(2024), 7 [11 pp.]$$pProc. ACM SIGGRAPH$$tProceedings of ACM SIGGRAPH$$x2376-1180 000147897 8564_ $$s970752$$uhttps://zaguan.unizar.es/record/147897/files/texto_completo.pdf$$yVersión publicada 000147897 8564_ $$s2512023$$uhttps://zaguan.unizar.es/record/147897/files/texto_completo.jpg?subformat=icon$$xicon$$yVersión publicada 000147897 909CO $$ooai:zaguan.unizar.es:147897$$particulos$$pdriver 000147897 951__ $$a2025-01-09-14:41:27 000147897 980__ $$aARTICLE