Text2Fit: An End-to-End Virtual Try-On and Poster Generation Agent Based on Diffusion Models
Abstract
Virtual try-on has achieved remarkable visual quality on benchmarks, yet most research focuses on generating isolated try-on images. In real e-commerce workflows, merchants need complete promotional assets including vertical posters with try-on previews, thumbnails, pricing, and QR codes, which still require extensive manual design. This paper presents Text2Fit, an end-to-end agent system that automatically produces both high-quality try-on images and mobile-ready e-commerce posters from person and garment images. Built upon PromptDresser, Text2Fit introduces three novel agent capabilities including an input analyzer that profiles image quality and estimates processing difficulty, a strategy planner that dynamically selects preprocessing and diffusion parameters, and an adaptive executor with quality assessment that generates multiple candidates and selects the best result using four complementary metrics including a garment texture preservation score specifically designed for e-commerce scenarios. We also introduce a layout-aware poster generator that composes try-on results with product metadata into consistent vertical posters. Experiments on VITON-HD demonstrate that Text2Fit achieves comparable try-on quality to PromptDresser (FID: 8.64, LPIPS: 0.1165, SSIM: 0.8530) while providing additional capabilities. User studies show that 85% of generated posters require minimal or no manual editing.
Read the paper
Related studies
- Learning from First-Generation Qualitative Approaches in the IS Discipline: An Evolutionary View and Some Implications for Authors and Evaluators (PART 1/2)
- Commentary on the clinical and preclinical dosage limits of interstitially administered magnetic fluids for therapeutic hyperthermia based on current practice and efficacy models
- Prosthesis: New Technological Opportunities and Innovative Biomedical Devices
- The history and future of neural modeling for cochlear implants
- Ryanodine receptors: waking up from refractoriness
- Novel Transgenic Rabbit Model Sheds Light on the Puzzling Role of Matrix Metalloproteinase-12 in Atherosclerosis
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.