Releases: marco-willi/camera-trap-classifier
Releases · marco-willi/camera-trap-classifier
New Trained Models / Support for Docker, Setuptools and more
New Features:
- new models for download (trained on over 3.6 million images)
- Xception architecture
- options to install/use the code: setuptools, command line tool, Docker, Singularity
- new data augmentation options: zoom, rotate, crop
- predict and aggregate multiple images of a capture together
Other:
- upgrade to Tensorflow 1.12
- various code clean-ups and small fixes
Major revision with improved usability
- streamline process into distinct parts
- command line only interface
- support of multi-image captures & missing values
- instructions on how to deploy models
- add link to Zenodo
Release for Zenodo publication
2.0.1 add link to readme
First Version after Panthera Implementation
This is the initial version after implementing the code for Panthera.