Bruce

原有的MN3好像写错了？

1

MN2 = np.concatenate([missing1[np.newaxis, :], (len(kp2)) * np.ones((1, len(missing1)), dtype=np.int64)]) MN3 = np.concatenate([missing2[np.newaxis, :], (len(kp1)) * np.ones((1, len(missing2)), dtype=np.int64)])


How to test in the nuscenses?

1

Excuse me, dear author! There are only train.py and benchmark.py for inference speed.

Excuse me , why the param training=True in eval? If training=True in eval, the mode in handshake will be softmax instead of argmax/activated_selected.

1

https://github.com/GT-RIPL/MultiAgentPerception/blob/4ef300547a7f7af2676a034f7cf742b009f57d99/ptsemseg/trainer.py#L968

https://github.com/GT-RIPL/MultiAgentPerception/blob/4ef300547a7f7af2676a034f7cf742b009f57d99/ptsemseg/trainer.py#L1028

How to visualize the predictions?

1

Excuse me, is there any available tool to visualize the predicted results?

How to generate dataset labels with the same format as nuscenes

4

CARLA version: Platform/OS: Ubuntu Problem you have experienced: How to generate dataset labels with the same format as nuscenes What you expected to happen: The related source code