search

CVLAB-Unibo / MM2D3D

Official code for the paper "Exploiting the Complementarity of 2D and 3D Networks to Address Domain-Shift in 3D Semantic Segmentation"
10 stars 0 forks source link

code #2

Open xin1106 opened 1 year ago

xin1106 commented 1 year ago

hi @adricarda ,

thank you for your work.

I don't see the processing code for the A2D2 data set and semantic-kitti data set. Could you share the preprocess_a2d2.py and preprocess_semantic-kitti.py ?

Thanks! 11W{P7W}$OLPU Q0E1`R{J6

adricarda commented 1 year ago

Hi @xin1106, thank you for your interest. The preprocessing code should be the same used in this repo: https://github.com/valeoai/xmuda_journal

thanks

xin1106 commented 1 year ago

Thank you for your reply, but I still have questions. In a2d2.py, we use the variable ‘pts_cam_coord’, but the data_dict generated by the preprocessing code (https://github.com/valeoai/xmuda_journal) does not contain the ‘pts_cam_coord‘.

a2d2.py pts_cam_coord = data_dict["pts_cam_coord"].copy()

preprocess.py (https://github.com/valeoai/xmuda_journal) data_dict['seg_label'] = seg_label.astype(np.uint8) data_dict['points'] = points.astype(np.float32) data_dict['points_img'] = points_img # row, col format, shape: (num_points, 2) data_dict['img'] = image

xin1106 commented 1 year ago

@adricarda Look forward to your kind reply！