README

What's new:
- run_evaluation_recognition.py
- config_recognition.json
- metrics/evaluation_recognition.py
- feature_extractors/
- data/perfectly_detected_ears/
- data/ears/annotations/recognition/ (class annotations are stored here)

FOR TESTING:
Use your own predictions from the previous task as the input for recognition - copy detected images into "data/<yourdetectorsname>" (like "data/perfectly_detected_ears"). If your detector perfomed poorly or you would like to start from scratch, feel free to use "data/perfectly_detected_ears" directly.

FOR TRAINING:
Use "data/perfectly_detected_ears" to train/prepare your model/approach.