For CNN image recognition tasks, like object recognition/face recognition/object segmentation/posture recognition, are there experiment results about how much will the performance be degraded with monochrome images?
The imaginary experiment is like:
Take the existing frameworks, reduce the channel number in the framework to fit monochrome images
Transform the existing training data and testing data to monochrome images
Train the model with monochrome training data.
Use the model to test the monochrome testing data.
Compare the result with the original result.