stevenwudi / 6dvnet Goto Github PK

❓ Questions and Help

Thank you for your work!
I used your code to run and found that there are quite a few deviations between the model and the real vehicle. There also are some problems about the vehicle model matching. I don't know if you have encountered this problem and have a solution?

❓ Questions and Help

I've loaded the pickle model for car instance segmentation and encountered the following error in apolloscape_car_3d.py:

line 190, in _add_gt_annotations_Car3d
car = self.car_models[car_name]
TypeError: 'CHJ_tiny_obj' object is not subscriptable

Is there any other way to extract the data from the file?
Also provide steps for training apollo dataset.

运行这句时
nvidia-docker build -t maskrcnn-benchmark --build-arg CUDA=10.0 --build-arg CUDNN=7 docker/
报错，请问该如何解决？

❓ Questions and Help

Thanks for your great work, but your codes are really in a bad order, and the docs you made are really rough, which is really hard for beginners to understand. So could you please update them to a better version?

❓ Questions and Help

on
6DVNET/maskrcnn_benchmark/data/datasets/apolloscape_car_3d.py

line 181

for face in car['faces'] - 1:
      pts = np.array([[imgpts[idx, 0], imgpts[idx, 1]] for idx in face], np.int32)
      pts = pts.reshape((-1, 1, 2))
      #cv2.polylines(mask, [pts], True, (255, 255, 255))
      cv2.drawContours(mask, [pts], 0, (255, 255, 255), -1)

❓ Questions and Help

这个代码里没有加论文里的Weighted Non-local neighbour embedding模块吧？

❓ Questions and Help

There are so many datasets include, e.g, Apollo3D, Pascal3D, and KITTI. Could you please provide a more detailed readme about how to use your code? Thanks in advance.