Wildlife images captured in a field represent a challenging task while classifying animals since they appear with a different pose, cluttered background, different light and climate conditions, different viewpoints, and occlusions. Additionally, animals of different classes look similar. All these challenges necessitate an efficient algorithm for classification.
This code was originally developed for this hackerearth challenge. The model shown is a normal CNN on augmented data.
The link for the dataset can be found here.
- Keras 2.2 with Tensorflow backend