AI RESEARCH × REAL-WORLD IMPACT
Building intelligence for the visual world.
FabricVTON researches and develops AI systems that understand, generate and transform visual reality — from people and products to materials and environments.
OUR RESEARCH
We work on problems where vision meets reality.
From fabric and humans to objects and environments, we build models that see deeper, understand better and generate with greater control.
Explore research
01Visual UnderstandingModels that understand people, objects, garments and environments.
02Generative VisionSystems that can transform visual inputs while preserving structure and identity.
03Material IntelligenceUnderstanding appearance, texture, geometry and physical characteristics.
04Human–Object InteractionModeling how objects change when interacting with people and in the real world.
OPEN PROBLEMS
What we’re working on now
These problems are hard and unsolved. We treat them as research, with experiments, ablations and benchmarks, and the results ship in production.
Read about the open problems Work on these with us- 01Fabric fidelity through warpingKeeping texture, print and logos intact while a garment is warped onto a body.
- 02Drape and foldsRealistic drape, folds and fit across different body shapes.
- 03Pose and camera consistencyThe same garment, consistent across poses and camera angles.
- 04Fast, affordable inferenceGeneration fast and cheap enough to run on a live storefront.
OUR APPROACH
More than pixels.
A deeper understanding.
We combine computer vision, generative models and physical understanding to build AI that respects the real world — its materials, geometry and complexity.
Our approach- InputImage
- GeometryStructure
- MaterialTexture & Fabric
- ModelGenerative AI
- OutputRealistic Result
Illustrative pipeline.
OUR PRODUCT
Research
in the real world.
Clothsy AI is our first product — an AI-powered virtual try-on platform for fashion commerce, built on top of FabricVTON’s research in visual understanding, material modeling and generative AI.
FROM OUR JOURNAL
Ideas, experiments and what we’re learning.
A behind-the-scenes look at our research, engineering and journey building visual AI.
View all posts
Why fabric is harder than pixels
Exploring the unique challenges of modeling texture, drape and material properties in generative AI.
Evaluating texture fidelity in VTO
A framework for measuring how well AI preserves print, weave and material identity.
Pose consistency in garment generation
Why one good frame isn’t enough, and how we maintain identity across views and poses.
COMPANY
We started with a hard problem.
Visual AI can generate remarkable images. But real-world understanding requires more than appearance alone.
FabricVTON was built around the idea that systems should understand structure, materials, people and the ways they interact.
We research the underlying technology and turn that work into products people can use.
Read our visionCAREERS
Work on problems that don’t have clean answers.
We are building systems for visual AI where texture, geometry, identity, realism and efficiency all matter at once.
We’re building a virtual try-on research team of students, researchers and engineers. The work is real research, and it ships in production. The application takes about two minutes.
LET’S BUILD A MORE VISUAL WORLD
Research. Product. People.
Join us in shaping the next generation of visual intelligence.