M. Uluç Şahin

face_ind.txt is used for getting vertices that correspond to face area. I was wondering how this was computed. I trained model with an extra deconv layer at the end for...

How would we go if we wanted to implement an evaluation metric for generator part? I have tried to load a pre-trained discriminator addition to the discriminator that trained during...

I was wondering if it is possible to see C++ implementation of Pytorch3dRasterizer.

Results

1

Hello, Does this implementation give same results as the original implementation? Thanks.

I want to visualize keypoints on images. However, keypoints obtained from KPDetector doesn't make sense to me as they are values around zero. What are the scaling and translation values...

Pretrained Model

5

Hello, I was wondering if you are going to share a pre-trained model? Also thanks for the implementation.

It is like 256x256 results are upscaled to 512x512. I set sigma to 1.5 and used 512x512 inputs. I removed resizing operations in the code. Is there anything else I...

Inference Speed

1

I trained a model with provided Wojtek data, inference speed is around 1.4 seconds per frame. I also trained a model using different data which I generated following the steps...

I was wondering the reasoning behind this change. As far as I know, batch_rodrigues method was being used in FLAME to compute rotation matrix. I also see that full_pose has...

Hello, How would you do inference on this model? Evaluator runs without a problem but I couldn't find how to make inference by feeding individual audio to model. Thanks.