yg33717 / DSH_tensorflow

implemement of DEEP SUPERVISED HASHING FOR FAST IMAGE RETRIEVAL_CVPR2016
105 stars 37 forks source link

Questions about pairwise loss in the hashing_loss #6

Open lcgrace opened 6 years ago

lcgrace commented 6 years ago

Thank you for your codes so much. But is it sure to implement paired input in hash_loss? The label in w_label = tf.matmul(label, label, False, True) is the same. w_label is the similarity of the n(n-1) image pairs?

world2025 commented 6 years ago

@lcgrace Hi, I also have some questions with the paired inputs and the loss. I think the loss is not in the papaer,do you understand and solve it ? If you get it,please tell me ,thank you very much.