search

jimmyyhwu / pose-interpreter-networks

Real-time robotic object pose estimation with deep learning
MIT License
122 stars 27 forks source link

How to use the pretrained object pose estimation model for end to end eval. #10

Closed yikakabu closed 5 years ago

yikakabu commented 5 years ago

Hi, Jimmy. When I loaded the checkpoint of floating_kinect1_object in the end_to_end_eval.ipynb. There is a error

RuntimeError: Given groups=1, weight[64, 3, 7, 7], so expected input[1, 1, 240, 320] to have 3 channels, but got 1 channels instead

So I wander could the object chekpoint be used to end_to_end_eval/visualize. Thanks you.

dingshenglan commented 5 years ago

I met this problem，just because the mask png has only one channel.

jimmyyhwu commented 5 years ago

No, the end-to-end expects a mask input, so you can't use floating_kinect1_object.

yikakabu commented 5 years ago

@jimmyyhwu I see. Thank you for your reply.