|
1 | 1 | # Changelog of v1.1
|
2 | 2 |
|
| 3 | +### v1.1.1 (30/5/2023) |
| 4 | + |
| 5 | +#### Highlights |
| 6 | + |
| 7 | +- Support [TPVFormer](https://arxiv.org/pdf/2302.07817.pdf) in `projects` (#2399, #2517, #2535) |
| 8 | +- Support the training of \[BEVFusion\] in `projects` (#2546) |
| 9 | +- Support lidar-based 3D semantic segmentation benchmark (#2530, #2559) |
| 10 | + |
| 11 | +#### New Features |
| 12 | + |
| 13 | +- Support [TPVFormer](https://arxiv.org/pdf/2302.07817.pdf) in `projects` (#2399, #2517, #2535) |
| 14 | +- Support the training of \[BEVFusion\] in `projects` (#2558) |
| 15 | +- Support lidar-based 3D Semantic Segmentation Benchmark (#2530, #2559) |
| 16 | +- Support test-time augmentation for Segmentor (#2382) |
| 17 | +- Support `Minkowski ConvModule` and `Residual` Block (#2528) |
| 18 | +- Support the visualization of multi-view images in multi-modal methods (#2453) |
| 19 | + |
| 20 | +#### Improvements |
| 21 | + |
| 22 | +- Upload checkpoints and training log of PETR (#2555) |
| 23 | +- Replace `np.float` by default `float` in segmentation evaluation (#2527) |
| 24 | +- Add docs of converting SemanticKITTI datasets (#2515) |
| 25 | +- Support different colors for different classes in visualization (#2500) |
| 26 | +- Support tensor-like operations for `BaseInstance3DBoxes` and `BasePoint` |
| 27 | +- Add information of LiDAR Segmentation in NuScenes annotation files |
| 28 | +- Provide annotation files of datasets generated offline (#2457) |
| 29 | +- Refactor document structure (#2429) |
| 30 | +- Complete typehints and docstring (#2396, #2457, #2468, #2464, #2485) |
| 31 | + |
| 32 | +#### Bug Fixes |
| 33 | + |
| 34 | +- Fix the bug of abnormal loss when training SECOND in Automatic mixed precision(AMP) mode (#2452) |
| 35 | +- Add a warning in function `post_process_coords` in mmdet3d/dataset/convert_utils.py (#2557) |
| 36 | +- Fix invalid configs (#2477, #2536) |
| 37 | +- Fix bugs of unit test (#2466) |
| 38 | +- Update `local-rank` argument in test.py for pytorch 2.0 (#2469) |
| 39 | +- Fix docker file (#2451) |
| 40 | +- Fix demo and visualization (#2453) |
| 41 | +- Fix SUN RGB-D data converter (#2440) |
| 42 | +- Fix readthedocs building (#2459, #2419, #2505, #2396) |
| 43 | +- Fix CI #(2445) |
| 44 | +- Fix the version error of `torch` in github merge stage test (#2424) |
| 45 | +- Loose the version restriction of `numba` (#2416) |
| 46 | + |
| 47 | +#### Contributors |
| 48 | + |
| 49 | +A total of 10 developers contributed to this release. |
| 50 | + |
| 51 | +@sunjiahao1999, @Xiangxu-0103, @JingweiZhang12, @chriscarving, @jaan1729, @pd-michaelstanley, @filaPro, @kabouzeid, @A-new-b, @lbin |
| 52 | + |
3 | 53 | ### v1.1.0 (6/4/2023)
|
4 | 54 |
|
5 | 55 | #### Highlights
|
|
0 commit comments