Khairul Imam comments

Repositories
Issues
Comments

Results 4 comments of


                                            Khairul Imam

Why use " labels.append(np.ones(dists.size(0)))" in training? and

That basically means that all predictions made by comparing `anc` and `neg` embeddings should by default labeled as 0 meaning not same. This is intuitively confusing since we get used...

Why use " labels.append(np.ones(dists.size(0)))" in training? and

> Hi, thank you for you nice job! I feel puzzled why there is only the 'triplet_loss' but no 'CrossEntropyLoss' during training. It seems that 'model.module.forward_classifier' here not be used...

Pretrained Model on vggface2

I've done the training on VGGv2, You can find the pretrained weight here https://github.com/khrlimam/res-facenet

Is there pretrained weights available?

I've done the training. You can find the weight here https://github.com/khrlimam/res-facenet