How to train the model on kitti dataset？

[406143996]

Thank the author for the great work!

I want to run through this model on the kitti format data set. After modifying the data part in the configuration file and the relevant part in the train_pipeline, I find it is always difficult to run through, and the following bugs appear. Is there a configuration file description for kitti dataset in this code?

Traceback (most recent call last): File "tools/train.py", line 249, in <module> main() File "tools/train.py", line 238, in main train_model( File "/public/chenrunze/wl/mm2/mmdetection3d-0.17.1/mmdet3d/apis/train.py", line 28, in train_model train_detector( File "/public/chenrunze/miniconda3/envs/di_help_zq/lib/python3.8/site-packages/mmdet/apis/train.py", line 170, in train_detector runner.run(data_loaders, cfg.workflow) File "/public/chenrunze/miniconda3/envs/di_help_zq/lib/python3.8/site-packages/mmcv/runner/epoch_based_runner.py", line 127, in run epoch_runner(data_loaders[i], **kwargs) File "/public/chenrunze/miniconda3/envs/di_help_zq/lib/python3.8/site-packages/mmcv/runner/epoch_based_runner.py", line 50, in train self.run_iter(data_batch, train_mode=True, **kwargs) File "/public/chenrunze/miniconda3/envs/di_help_zq/lib/python3.8/site-packages/mmcv/runner/epoch_based_runner.py", line 29, in run_iter outputs = self.model.train_step(data_batch, self.optimizer, File "/public/chenrunze/miniconda3/envs/di_help_zq/lib/python3.8/site-packages/mmcv/parallel/distributed.py", line 52, in train_step output = self.module.train_step(*inputs[0], **kwargs[0]) File "/public/chenrunze/miniconda3/envs/di_help_zq/lib/python3.8/site-packages/mmdet/models/detectors/base.py", line 237, in train_step losses = self(**data) File "/public/chenrunze/miniconda3/envs/di_help_zq/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1051, in _call_impl return forward_call(*input, **kwargs) File "/public/chenrunze/miniconda3/envs/di_help_zq/lib/python3.8/site-packages/mmcv/runner/fp16_utils.py", line 98, in new_func return old_func(*args, **kwargs) File "/public/chenrunze/wl/mm2/mmdetection3d-0.17.1/mmdet3d/models/detectors/base.py", line 59, in forward return self.forward_train(**kwargs) File "/public/chenrunze/zq/DeepInteraction-main/projects/mmdet3d_plugin/models/detectors/deepinteraction.py", line 200, in forward_train img_feats, pts_feats = self.extract_feat( File "/public/chenrunze/zq/DeepInteraction-main/projects/mmdet3d_plugin/models/detectors/deepinteraction.py", line 140, in extract_feat new_img_feat, new_pts_feat = self.imgpts_neck(img_feats[0], pts_feats[0], img_metas, pts_metas) TypeError: 'NoneType' object is not subscriptable

[xtk8532704]

There are some modules in the code that depend on the nuscenes dataset. You need to read the code carefully and modify them.

[HydrogenWasser]

@406143996 hi, have u trained deepinteraction on kitti? i modified it to suit kitti dataset but the training result is sooo bad, i wonder if i made some mistakes.

[enestydarealmc]

@406143996 hi, have u trained deepinteraction on kitti? i modified it to suit kitti dataset but the training result is sooo bad, i wonder if i made some mistakes.

can you please share your config? I'd be very grateful. Thank you

[HydrogenWasser]

@406143996 hi, have u trained deepinteraction on kitti? i modified it to suit kitti dataset but the training result is sooo bad, i wonder if i made some mistakes.

can you please share your config? I'd be very grateful. Thank you

same bad result, cant train deepinteraction on kitti