Suvi-dha

Thank you. Could you please help me how I could get 3d landmarks from the scan without manual extraction?

Yes, its same as pncc that is converting normalized x,y,z values to r,g,b values. The difference is that here x, y, z coordinates in 3d space is mapped to texture...

downgrade scipy to scipy 1.8.1

Even I am facing this now, training on 65M images the loss has stopped converging after epoch 0.

I want to add gradient vector flow (external force)

You can use this code to generate landmark.txt ```` import face_alignment import cv2 import numpy as np fa = face_alignment.FaceAlignment(face_alignment.LandmarksType._2D, flip_input=False) img = cv2.imread('./test.png') preds = fa.get_landmarks(img)[0] # (68,2) w0...

This worked for me: 1. uninstall ffmpeg-python 2. pip install ffmpeg 3. pip install ffmpeg-python

I am getting this error when training to train just shape maps using your pretrained network. And I don't have labels in my training data. `Traceback (most recent call last):...

Hi @rohit7044 I have few 3D models with their textures. I converted those 3D models to shape maps and normals.

Hi, have you tried with latest versions