Finding Waldo (with YOLOv8)


Total training images: 219
Epochs: 30
Model size: 136.8MB
Precision: 0.87 (how accurate predictions are; higher = better)
Recall: 0.876 (how good it is at finding positives; higher = better)
mean Average Precision (mAP@[.5:.95]): 0.638 (mAP over different IoU thresholds; higher = better)
Total training time: 33min (Vertex AI Workbench Notebook with NVIDIA Tesla T4 GPU)

Select a file to upload


- Uploaded images are automatically deleted after 24 hours
- Portrait photos from mobile recommended (landscape photos submitted as portraits and may not work as expected)