← All Projects
Computer Vision hard ~45 hours
Multi-Scale Defect Detection with SAHI
Build an industrial defect detection system that handles tiny defects using SAHI (tiled inference), FPN-based detection, and a custom evaluation pipeline that tracks detection at each scale.

Skills Demonstrated

Object detection (YOLOv8/RT-DETR) SAHI tiled inference Multi-scale evaluation Production deployment with TensorRT

Implementation Steps

  1. Collect/curate defect dataset with multi-scale annotations
  2. Train YOLOv8 baseline and measure small-object recall
  3. Implement SAHI tiled inference pipeline
  4. Add FPN-based detection head for tiny objects
  5. Build per-scale evaluation metrics (mAP@small, medium, large)
  6. Export to TensorRT for production-speed inference
  7. Create Gradio demo with live camera feed

Interview Relevance

Why this project matters for interviews Manufacturing AI and visual inspection are massive markets. Showing you can handle the multi-scale challenge with real metrics is directly relevant to roles at Tesla, Apple Vision, and startups.
All Projects Back to Interview Prep