Mission accomplished
The mission was to provide an objective evaluation of the fruit classification model for the convenience store chain. The predictive performance on out-of-sample validation images was dismal! You could have stopped there, but then you would not have known how to make a better model.
However, the predictive performance evaluation was instrumental in deriving specific misclassifications, as well as correct classifications, to assess using other interpretation methods. To this end, you ran a comprehensive suite of interpretation methods, including activation, gradient, perturbation, and backpropagation-based methods. The consensus between all the methods was that the model was having the following issues:
- Differentiating between the background and the fruit
- Understanding that different fruit classes share some color hues
- Confounding lighting conditions such as specular highlights and shadows as specific fruit characteristics
- Being confused...
