Person Re-identification is the task of matching images of a person across multiple camera views.
This is the implementation of the paper End-to-End Comparative Attentional Networks for person reid in torch.
1.A very nice tutorial for generative vs discriminative modelling.
2.Some Papers on deep learning.
3.A couple of blogs on RNN: R2R and WildML. These blogs have explanation of each and everything related to RNN in a very nice manner.
2.A video on Google's inception architecture.