## Rules

This competition is designed to be fair and learning-focused.

### ✅ Allowed
- Using only the dataset provided in this competition
- Training your own model from scratch
- Standard augmentations (flip, crop, color jitter, etc.)
- Pretrained architectures **only if trained on ImageNet** (example: ResNet, EfficientNet, ViT pretrained on ImageNet)

### ❌ Not Allowed
- Using any external dataset (anime dataset, face dataset, etc.)
- Using any external pretrained model trained on anime datasets
- Using any model or weights trained using external anime/human gender datasets

🚫 **No external datasets or anime-special pretrained weights allowed.**

All submissions may be checked. If external data/models are used, the submission may be disqualified.

This is mainly a fun competition — please play fair ❤️