Skip to content

SnakeSense: Snake Species Classifier Initial Release

Latest
Latest

Choose a tag to compare

View all tags
@lostplusfound lostplusfound released this 24 May 01:40
· 3 commits to main since this release
v1.0.0
2976fe7
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

SnakeSense Release v1.0

This release introduces SnakeSense, a snake species classifier AI model available in two formats:

  • ONNX model (model.onnx) — optimized for cross-platform deployment, outputs a predicted species index (0–134).
  • FastAI PyTorch model (model.pkl) — convenient for direct use with FastAI, outputs a predicted class ID.

Model Details

  • Architecture: ResNet101
  • Framework: FastAI
  • Training Platform: Kaggle
  • Dataset: 165 Different Snakes Species
  • Training Epochs: 100
  • Training Accuracy: 66.1%

Model Outputs & Lookup

Format Output Lookup Key in species.csv
.pkl (FastAI) class_id (integer) class_id
.onnx index (integer) index

Use the included species.csv file to map model outputs to detailed species information including common name, scientific name, venom status, and geographic data.

Included Files

  • model.onnx — ONNX export of the trained model
  • model.pkl — FastAI exported model for easy inference in Python
  • species.csv — CSV mapping indices and class IDs to species metadata

Usage Examples

  • Use model.pkl with FastAI for simple Python inference and easy integration.
  • Use model.onnx for cross-platform inference with ONNX Runtime.

Refer to the README for detailed code examples.

Disclaimer

This is an early version intended for experimentation. The model is not production-ready, and no AI can guarantee perfect accuracy. Please seek professional help if bitten by a snake and do not rely solely on this tool for safety decisions.

Feedback, contributions, and questions are highly welcome!

Assets 5
Loading