@@ -12,49 +12,28 @@ ______________________________________________________________________
12
12
13
13
## Install mmdet3d
14
14
15
- These branches are required for mmdet3d deployment
16
-
17
- | codebase | commit |
18
- | :------: | :-------: |
19
- | mmdet3d | v1.1.0rc1 |
20
- | mmcv | v2.0.0rc1 |
21
- | mmdet | v3.0.0rc1 |
22
- | mmseg | v1.0.0rc0 |
23
-
24
- First checkout and install mmcv/mmdet/mmseg/mmdet3d
25
-
26
- ``` bash
27
- python3 -m pip install openmim --user
28
- python3 -m mim install mmcv==2.0.0rc1 mmdet==3.0.0rc1 mmseg==1.0.0rc0 --user
29
-
30
- git clone https://github.com/open-mmlab/mmdetection3d --branch v1.1.0rc1
31
- cd mmdetection3d
32
- python3 -m pip install .
33
- cd -
34
- ```
35
-
36
- After installation, ` tools/check_env.py ` should display mmdet3d version normally
15
+ We could install mmdet3d through [ mim] ( https://github.com/open-mmlab/mim ) .
16
+ For other ways of installation, please refer to [ here] ( https://mmdetection3d.readthedocs.io/en/latest/get_started.html#installation )
37
17
38
18
``` bash
39
- python3 tools/check_env.py
40
- ..
41
- 11/11 13:56:19 - mmengine - INFO - ********** Codebase information**********
42
- 11/11 13:56:19 - mmengine - INFO - mmdet: 3.0.0rc1
43
- 11/11 13:56:19 - mmengine - INFO - mmseg: 1.0.0rc0
44
- ..
45
- 11/11 13:56:19 - mmengine - INFO - mmdet3d: 1.1.0rc1
19
+ python3 -m pip install -U openmim
20
+ python3 -m mim install " mmdet3d>=1.1.0"
46
21
```
47
22
48
23
## Convert model
49
24
50
25
For example, use ` tools/deploy.py ` to convert centerpoint to onnxruntime format
51
26
52
27
``` bash
53
- export MODEL_CONFIG=/path/to/mmdetection3d/configs/centerpoint/centerpoint_pillar02_second_secfpn_head-circlenms_8xb4-cyclic-20e_nus-3d.py
28
+ # cd to mmdeploy root directory
29
+ # download config and model
30
+ mim download mmdet3d --config centerpoint_pillar02_second_secfpn_head-circlenms_8xb4-cyclic-20e_nus-3d --dest .
31
+
32
+ export MODEL_CONFIG=centerpoint_pillar02_second_secfpn_head-circlenms_8xb4-cyclic-20e_nus-3d.py
54
33
55
- export MODEL_PATH=https://download.openmmlab.com/mmdetection3d/v1.0.0_models/centerpoint/centerpoint_02pillar_second_secfpn_circlenms_4x8_cyclic_20e_nus/centerpoint_02pillar_second_secfpn_circlenms_4x8_cyclic_20e_nus_20210816_064624-0f3299c0 .pth
34
+ export MODEL_PATH=centerpoint_02pillar_second_secfpn_circlenms_4x8_cyclic_20e_nus_20220811_031844-191a3822 .pth
56
35
57
- export TEST_DATA=/path/to/mmdetection3d/ tests/data/nuscenes/sweeps/LIDAR_TOP/ n008-2018-09-18-12-07-26-0400__LIDAR_TOP__1537287083900561 .pcd.bin
36
+ export TEST_DATA=tests/data/n008-2018-08-01-15-16-36-0400__LIDAR_TOP__1533151612397179 .pcd.bin
58
37
59
38
python3 tools/deploy.py configs/mmdet3d/voxel-detection/voxel-detection_onnxruntime_dynamic.py $MODEL_CONFIG $MODEL_PATH $TEST_DATA --work-dir centerpoint
60
39
```
@@ -82,4 +61,4 @@ The caller needs to refer to the corresponding [python implementation](../../../
82
61
| [ pointpillars] ( https://github.com/open-mmlab/mmdetection3d/blob/main/configs/pointpillars/pointpillars_hv_secfpn_8xb6-160e_kitti-3d-3class.py ) | voxel detection | KITTI | ✔️ | ✔️ | ✔️ |
83
62
| [ smoke] ( https://github.com/open-mmlab/mmdetection3d/blob/main/configs/smoke/smoke_dla34_dlaneck_gn-all_4xb8-6x_kitti-mono3d.py ) | monocular detection | KITTI | ✔️ | x | ✔️ |
84
63
85
- - Make sure trt >= 8.4 for some bug fixed, such as ScatterND, dynamic shape crash and so on.
64
+ - Make sure trt >= 8.6 for some bug fixed, such as ScatterND, dynamic shape crash and so on.
0 commit comments