Virtual Try-on
Industry
Fashion
Service
AI & MLOps
Client’s Location
Hong Kong
Business need
- Blur human faces in photos users upload to the system.
- Uses Virtual Try-on technology to match the manufacturer’s clothing to the baby photo the user uploads and deblur the human’s face.
AI Solution
- Detect the position of the face in the input image and blur the image using Gaussian Blur.
- Detect points in the input image and costume points using skeletal detection.
- Distinguish between costume parts and human parts in an input image using human parsing.
- Virtual try-on models use costumes that overlap the input image figure.
- Deblur the image and send it to the user.
Tech Stack
- Languages and Frameworks: Python 3, OpenCV, PyTorch, TensorFlow
- AI Techniques: Pose Estimation, Image Segmentation, Generative Adversarial Network（GAN）
- Cloud Computing: Microsoft Azure
Let’s make your project a success!
Partnering with AVEO isn’t just about digital transformation — it’s about strategic collaboration, expert guidance, and leveraging the full potential of technology to drive real business results. We don’t just build solutions; we help you turn them into measurable growth.