random resuls of landmark model (-1)

[sasanasadiabadi]

Trying the new landmark model, It seems that I get random landmarks. You mention that accuracy is very low, but I wonder if I'm doing some mistake! could you share some sample results of the new landmark model as well? thanks.

Also I think the reshaping in here

OpenSeeFace/model.py

Line 210 in 8ff2e61

t_off_x = x[30:60].reshape((60, 7*7)).gather(1, indices).squeeze(1)

OpenSeeFace/model.py

Line 211 in 8ff2e61

t_off_y = x[60:90].reshape((60, 7*7)).gather(1, indices).squeeze(1)

should be (30, 7*7).

[emilianavt]

You are right, that should be reshape(30, 7*7). I actually forgot to test the inference=True version, because tracker.py is currently doing the manual landmark decoding. Actually, I think I forgot to consider the batch size as well, even though I'm pretty sure I had it working correctly for the bigger models. I'll check what's going on there.

As a side note, the points for the 30 point model are laid out like this.

[emilianavt]

model.py should be fixed now in cfe2e2f. The models should work with bigger batch sizes and inference=True as well. The random output was most likely related to the different factor applied to the sigmoid to work better with the lower resolution. I've attached the output of the new model on one of the WIDER FACE images.

[sasanasadiabadi]

Thanks, the issue is fixed.

Before closing this, I'd like to ask if you still have the trained ShuffleNet-V2 PyTorch weights to share? Thanks.

[emilianavt]

I think I can find those, but they were trained on a less refined dataset and with less augmentation, so they're not really comparable. Are you still interested?

[sasanasadiabadi]

Yes that would be great if you could find them. I'm planning to retrain it, that'd be good to start from a pretrained model. Many thanks.

[emilianavt]

I think it should be these.

[sasanasadiabadi]

Thanks a lot. closing this issue.