Commits · f56e4f173ef4ce1da5ec3bed837d4175f96bb0ba · OpenDAS / OpenPCDet

02 Sep, 2022 5 commits
- remove self.seq_name_to_infos of WOD if we do not use multi-frame training · f56e4f17
  Shaoshuai Shi authored Sep 02, 2022
  
  f56e4f17
- add config of PV-RCNN++ 2frames · 72d587d4
  Shaoshuai Shi authored Sep 02, 2022
  
  72d587d4
- support to train single-frame model with multi-frame GT database on WOD · 7c85adb5
  Shaoshuai Shi authored Sep 02, 2022
  
  7c85adb5
- update codes and documents to update OpenPCDet to v0.6 · edb82b9f
  Shaoshuai Shi authored Sep 02, 2022
  
  edb82b9f
- merge with master branch · 1a42e442
  Shaoshuai Shi authored Sep 02, 2022
  
  1a42e442
23 Aug, 2022 3 commits
- update the function of random_world_translation · 58bca13f
  Shaoshuai Shi authored Aug 24, 2022
  
  58bca13f
- support to generate WOD gt_database with multi-frame tail · 8bf333f9
  Shaoshuai Shi authored Aug 23, 2022
  
  8bf333f9
- small bug in waymo_eval.py · 60e019e3
  Shaoshuai Shi authored Aug 23, 2022
  
  60e019e3
22 Aug, 2022 7 commits
- bugfixed (recently PR): eval_one_epoch args (#1074) · 7e8bbe26
  Shaoshuai Shi authored Aug 22, 2022
  
  7e8bbe26
- Update CUSTOM_DATASET_TUTORIAL.md · 3f70cbc7
  Shaoshuai Shi authored Aug 22, 2022
  
  3f70cbc7
- Support custom dataset template and tutorial - Merge pull request #1072 from jihanyang/master · b71c6a11
  Shaoshuai Shi authored Aug 22, 2022
```
Support custom dataset template and tutorial
```
  b71c6a11
- Merge branch 'dev_custom' into jh-master · d1e6cf04
  jihanyang authored Aug 22, 2022
  
  d1e6cf04
- Use category name in label format; Update README · e8bcd689
  jihanyang authored Aug 22, 2022
  
  e8bcd689
- add another custom dataset model config · 3cf5763f
  jihanyang authored Aug 22, 2022
  
  3cf5763f
- Modify the custom dataset support · cb1ecefa
  jihanyang authored Aug 22, 2022
  
  cb1ecefa
21 Aug, 2022 1 commit
- support to show gpu utilization/memory when training · 25d9f503
  Shaoshuai Shi authored Aug 21, 2022
  
  25d9f503
20 Aug, 2022 7 commits
- check np.float32/np.float64 when using create_integrated_database.py · d35088b6
  Shaoshuai Shi authored Aug 20, 2022
  
  d35088b6
- database should save with np.float32 · 96ba76a3
  Shaoshuai Shi authored Aug 20, 2022
  
  96ba76a3
- Merge branch 'master' into dev_v0.6 · 7e2d56b4
  Shaoshuai Shi authored Aug 20, 2022
  
  7e2d56b4
- Add process tools for create interated database; Support Inference latency calculation · a41e331d
  Shaoshuai Shi authored Aug 20, 2022
```
Add process tools for create interated database; Support Inference latency calculation
```
  a41e331d
- support to USE_CENTER_TO_FILTER in mask_points_and_boxes_outside_range · 5e21e428
  Shaoshuai Shi authored Aug 20, 2022
  
  5e21e428
- support inference latency when evaluation · ae26b88d
  jihanyang authored Aug 20, 2022
  
  ae26b88d
- Add a tool to create interated database from existing database; Add minor modifications to docs · bd405472
  jihanyang authored Aug 20, 2022
  
  bd405472
18 Aug, 2022 4 commits
- use p.map instead of p.imap for generating WOD gt_database · 6bd8be71
  Shaoshuai Shi authored Aug 18, 2022
  
  6bd8be71
- support to use multithreads (cpu/cuda) to generate gt_database of WOD · 38474d6c
  Shaoshuai Shi authored Aug 18, 2022
  
  38474d6c
- bugfixed: set empty gt_boxes_lidar to (0, 9) instead of (0, 7) · 6a1f253a
  Shaoshuai Shi authored Aug 18, 2022
  
  6a1f253a
- bugfixed: waymo_eval bug when training with speed · 4996eb46
  Shaoshuai Shi authored Aug 18, 2022
  
  4996eb46
17 Aug, 2022 9 commits
- support to train WOD with speed, support to update infos with speed · 1541d269
  Shaoshuai Shi authored Aug 17, 2022
  
  1541d269
- bugfixed to process waymo data · 25cd01e0
  Shaoshuai Shi authored Aug 17, 2022
  
  25cd01e0
- support MultiFrame setting for Waymo Open Dataset, add its config, needs to... · 505e0ab3
  Shaoshuai Shi authored Aug 17, 2022
```
support MultiFrame setting for Waymo Open Dataset, add its config, needs to re-preprocess the Waymo Dataset to save speeds/accels, save new GT database
```
  505e0ab3
- Default setting: DO NOT mask the gt_boxes that are outside the range of Z-axis · 053735e1
  Shaoshuai Shi authored Aug 17, 2022
  
  053735e1
- merge with master to support latest timer · 08652f94
  Shaoshuai Shi authored Aug 17, 2022
  
  08652f94
- support to use logger to store intermediate losses/states during training · 97a4e42e
  Shaoshuai Shi authored Aug 17, 2022
  
  97a4e42e
- Modify the timing manner of data timer and forward timer (#1064) · dd2edf9d
  jihan.yang authored Aug 17, 2022
  
  dd2edf9d
- Modify the timing manner of data timer and forward timer · 53cbe187
  jihanyang authored Aug 17, 2022
  
  53cbe187
- small update to README.md · 57b19553
  Shaoshuai Shi authored Aug 17, 2022
  
  57b19553
13 Aug, 2022 1 commit

feat: support torch>=1.11 (#1041) · 9ff29e24

Liu Yihua authored Aug 13, 2022

* feat: support torch>=1.11

Fix #900.
Support PyTorch version >= 1.11. Referring to https://github.com/pytorch/pytorch/pull/66765 and https://github.com/pytorch/pytorch/wiki/TH-to-ATen-porting-guide.

* fix: Remove preproc torch version check macros

9ff29e24

19 Jul, 2022 1 commit
- To make different spconv version compatible for focalsconv (#1034) · c233477a
  jihan.yang authored Jul 19, 2022
```
* To make different spconv version compatible for focalsconv
```
  c233477a
17 Jul, 2022 2 commits
- Merge pull request #1 from OrangeSodahub/custom_dataset · daa27fd8
  YangXiuyu authored Jul 17, 2022
```
Custom dataset
```
  daa27fd8
- add config files and README · e9b06b36
  YangXiuyu authored Jul 17, 2022
  
  e9b06b36