Photo quality first
The system looks for a clear full-body frame, visible shoulder-waist-hip outline, useful lighting, neutral posture, and clothing that does not hide the silhouette.
Methodology
How Body Shape Detector analyzes a photo.
The app is designed for clothing silhouette guidance only. It does not measure your body, identify you, or make health, attractiveness, identity, or protected-trait claims.
Silhouette proportions only
The prompt asks OpenAI to consider relative visual balance in the clothing silhouette. It is not allowed to estimate weight, BMI, age, ethnicity, sex, gender identity, disability, pregnancy, or other sensitive traits.
Match strength and caveats
Results use tentative, moderate, or strong match labels. Match strength drops when lighting, bulky clothing, pose, mirror angle, crop, or occlusion makes the silhouette ambiguous.
Practical styling output
The output focuses on clothing ideas, not body judgment. Recommendations are meant to help users test shapes, structure, drape, and proportions in real outfits.
Safety limits
Unknown is a valid result.
If the image is unsuitable or ambiguous, the app returns unknown and explains how to retake a better photo. That protects users from overly certain results when the image does not support one.
Try a new photo