CtrlVTON Enables Controllable Virtual Try-On
NXN-Labs · hf · 2026-07-14
This work introduces two related contributions:
- VIP-SAM (Visual-Instance-Prompt Segmentation)
- Aims to segment the exact same clothing instance between flat garment images and images of people wearing them.
- This is an instance-level task, distinct from common category-level segmentation.
- CtrlVTON
- Reformulates virtual try-on as an image editing problem.
- Incorporates segmentation masks to provide pixel-level control over the garment's fit, size, and position on the person (e.g., loose/fitted, tucked in/untucked, open/closed).
The authors claim both methods achieve SOTA in their respective tasks. Notably, CtrlVTON outperforms even the strongest commercial editing systems in adhering to user layouts while perfectly preserving garment details.
More from Multimodal
- Reddit user seeks ComfyUI NSFW text-to-image and image-to-video workflows under 20 GB VRAM — hobbyist2020 · 2026-07-22
- Krea 2 users recommend a two-pass Clownshark sampler setup for sharper image details — listopalafoto · 2026-07-22
- Gemini Omni Flash turns a boat cabin into a cave in Flow by Google — chrisfirst · 2026-07-22
- A simple workflow to turn a photo into an image prompt using Gemini, Grok, or GPT Image — harshitagu72595 · 2026-07-22
- A Reddit user proposes a consistency LoRA to keep anime and game scenes visually stable — ThirdWorldBoy21 · 2026-07-22
- Hand-painted figurines run through Seedance look eerily alive — cocktailpeanut · 2026-07-22