Yuzhi Liu, Enis Simsar, Giulia Lanzillotta, Thomas Hofmann · 2026 · arXiv
DOI: 10.48550/arXiv.2610.28567 · open-access PDF
La personalización de texto a imagen tiene como objetivo equipar modelos de difusión pre-entrenados con la capacidad de aprender conceptos específicos de los usuarios a partir de sólo unas pocas imágenes de referencia. Sin embargo, los métodos existentes suelen entrenar cada concepto de forma independiente o asumir que todos los conceptos están disponibles simultáneamente. En aplicaciones prácticas, los conceptos a menudo llegan secuencialmente, requiriendo que los modelos aprendan continuamente nuevos temas, objetos o estilos al tiempo que minimizan la interferencia con conceptos adquiridos previamente.
Machine translation (Meta M2M100); the English original follows.
Text-to-image personalization aims to equip pretrained diffusion models with the ability to learn user-specific concepts from only a few reference images. Existing methods, however, typically train each concept independently or assume that all concepts are available simultaneously. In practical applications, concepts often arrive sequentially, requiring models to continually learn new subjects, objects, or styles while minimizing interference with previously acquired concepts. We address this continual concept personalization setting by extending residual-space adaptation to the stochastic ada
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.28567, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange