EfficientNet-B2 Emotion Detection (v1)
This model classifies facial emotions into 7 categories:
angry, disgust, fear, happy, neutral, sad, surprise.
Architecture: EfficientNet-B2
Training Platform: Kaggle GPU
Accuracy: ~80.25%
Framework: PyTorch
Developer: Varad V. Choudhari (Atman AI)
License: MIT
Example Usage
from huggingface_hub import hf_hub_download
path = hf_hub_download(
repo_id="AtmanAI/emotion-detection-efficientnet-b2-v1",
filename="efficientnet_b2_emotion_final.pth"
)
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
Evaluation results
- accuracy on Facial Emotion Dataset (Kaggle)self-reported0.802