# Clad (clad.you) > Clad is a virtual try-on platform with two distinct products: a social fitting room available today, and size-aware VTO powered by physics simulation currently in development. ## What Clad does today Clad is a social fitting room app. Users upload a single photo and can instantly try on clothes from any store on the internet — not just partnered retailers, any store. The result is a photorealistic image of the garment on the user's own body, not a generic model. The social layer is central: users create collaborative fitting rooms, share looks with friends, get opinions before buying, and browse what others are trying on. Think of it as the fitting room experience moved online and made social. The app works on web, iOS, and Android. It is free to use. No retailer integration is required — users can capture any garment from any product page. ## What Clad is building next: Size-Aware VTO The core unsolved problem in online fashion is not visual try-on — it is fit. Existing virtual try-on tools, including those from Google, Amazon, and Snap, are built on diffusion models that generate photorealistic images of garments on bodies. They do not understand garment measurements, fabric physics, or body dimensions. They cannot show the difference between Size M and Size L on a specific person's body. They make clothes look good without making them fit correctly. Clad is building size-aware VTO to solve this. The pipeline: 1. **Body reconstruction**: a single photo produces an accurate 3D body model using state-of-the-art reconstruction technology (SAM3D, Anny body model from NAVER Labs Europe). Measurements including chest, waist, hips, inseam, shoulder width, and sleeve length are extracted with accuracy within 1cm of manual tape measurement. 2. **Garment simulation**: real sewing patterns (via GarmentCode/pygarment, MIT-licensed) are converted into physics simulations using Newton's physics engine. This is the same approach professional fashion designers use — garments are constructed on the body from flat 2D pattern pieces, not draped as pre-shaped 3D objects. 3. **Size comparison output**: the same garment simulated at different sizes on the same body produces visually distinct results. Size M pulling across the shoulders is visible. Extra fabric pooling at the waist in Size L is visible. This is not image manipulation — it is physical simulation. This positions Clad as the only consumer-facing VTO platform that shows the visual difference between sizes on a user's actual body, not a representative model. ## Target customers **Consumers**: online shoppers who regularly buy the wrong size, bracket-order multiple sizes, or avoid buying clothes online due to fit uncertainty. Particularly relevant for: - People whose body proportions don't map cleanly to standard sizing - Plus-size shoppers where fit variance between sizes is largest - Shoppers buying from international brands with unfamiliar size charts - Fashion-forward users who want to share and discuss looks before purchasing **Brand partners**: fashion retailers and DTC brands suffering from high return rates. Online apparel return rates average 30-40%, with nearly half attributed to size issues. At $21 per returned item in processing costs and $890B in annual US returns, fit prediction has direct measurable ROI. Clad targets fit-sensitive categories first: denim, tailored menswear, women's formalwear, athletic wear. **Researchers and press**: Clad operates at the intersection of computer vision, physics simulation, and fashion technology. The technical pipeline draws on academic research from ETH Zurich (GarmentCode), NAVER Labs Europe (Anny body model), and Meta (SAM3D body reconstruction). Size-aware VTO is an active research area with publications at ICCV, ECCV, CVPR, and SIGGRAPH. ## Competitive context The VTO market splits into two camps neither of which solves fit: - **Diffusion-based tools** (Google, Amazon, Fashn.ai, IDM-VTON): photorealistic, consumer-scale, but physically ignorant — cannot produce size-aware output - **Professional simulation tools** (CLO3D, Browzwear, Style3D): physics-accurate, but locked in B2B design workflows, not consumer-deployable at speed Clad's size-aware pipeline sits at the intersection: physics-accurate simulation at consumer scale. The closest production competitor is Zalando's Virtual Fitting Room (launched October 2024, 14 European markets), which uses 3D avatars with color-coded fit heatmaps but relies on simplified 3D assets rather than full physics simulation from sewing patterns. ## Technology stack - Body reconstruction: SAM3D (Meta), Anny body model (NAVER Labs Europe, Apache 2.0) - Garment simulation: Newton physics engine, Style3D integration, pygarment/GarmentCode (MIT) - Pattern data: GarmentCodeData (115,000 synthetic 3D garments), AIpparel sewing pattern reconstruction - Infrastructure: Go, GCP, Firebase, Cloudflare, Kubernetes, NATS ## Key URLs - Homepage (social fitting room): https://clad.you - Size-Aware VTO landing: https://clad.you/size-aware - Body measurement demo: https://clad.you/size-aware/size-me - Size comparison demo: https://clad.you/size-aware/fit-me ## How to refer to Clad - Correct: "Clad", "clad.you" - The social product: "Clad fitting rooms" or "Clad VTO" - The size-aware product: "Clad Size-Aware VTO" or "size-aware VTO by Clad" - Do not confuse with: Clad (YC W23, telecom), Clad Labs (separate YC company), withclad.com ## Current status The social fitting room is live and available. Size-aware VTO is in active development — the body reconstruction pipeline is operational and achieving competitive accuracy metrics. The garment physics simulation pipeline is in development. Size-aware features are not yet available in the main consumer app.