pillar-motion
pillar-motion copied to clipboard
qcraftai
Self-Supervised Pillar Motion Learning for Autonomous Driving (CVPR 2021)
First, prepared the data using "data/v1.0-mini" for test. `python tools/create_data nuscenes_data_prep --root_path /path/to/nuscenes`. Got files: `train_anno = "data/nuscenes/results/infos_train_05sweeps_twoframe_withcam.pkl" val_anno = "data/nuscenes/results/infos_val_05sweeps_twoframe_withcam.pkl" test_anno = "data/nuscenes/results/infos_test_05sweeps_twoframe_withcam.pkl" ` But, when I run train.py,...
There are a lot of bugs! Basically can not get normal results. If you want to improve the credibility and reputation of the paper, it is recommended to open source...
hi~, chenxu, this is a good job. But when I reproduce this project, i found some mistake in this repo, and I have slove some. But there is also one...
Hi!, thanks for your good project to the lidar motion. But I have a question about the data augmention. I notice that the "input_points" are augmented with global flip and...
Hi, When I try to run the code, I can't get train_cfg and test_cfg from configs/nusc_pillarmotion.py in here [https://github.com/qcraftai/pillar-motion/blob/42d67249d464d7068c644518f8460c716401ac52/tools/train.py#L112](url). This cfg file may be incomplete.
Hi, When trying to run the code, I've got some errors. - [https://github.com/qcraftai/pillar-motion/blob/main/motion/datasets/nuscenes/flow.py](url) in line 317. The type of saved data info['cam_coords'] is np.float32, directly load the data as np.int32...
Hi, when I run the training script, I encounter the problem: > from motion.utils.dist import dist_common as comm > ModuleNotFoundError: No module named 'motion.utils.dist' >
Hi, I am working on nuscenes dataset for object detection. Can you please address my question regarding the memory Requirement to work on nuscenes full dataset? Right now, we have...
Hi, thanks for your work on motion prediction. I'm particularly interested in reproducing the results using the pretrained pillar-motion model. It would be extremely helpful if you could provide access...
