Skip to content

Commit 5d6ed5d

Browse files
committed
refactor(lidar_centerpoint): update docs.
Signed-off-by: Kaan Çolak <kaancolak95@gmail.com>
1 parent 64f2958 commit 5d6ed5d

File tree

1 file changed

+6
-6
lines changed

1 file changed

+6
-6
lines changed

perception/lidar_centerpoint/README.md

+6-6
Original file line numberDiff line numberDiff line change
@@ -107,7 +107,7 @@ conda install pytorch==1.13.1 torchvision==0.14.1 pytorch-cuda=11.6 -c pytorch -
107107
pip install -U openmim
108108
mim install mmengine
109109
mim install 'mmcv>=2.0.0rc4'
110-
mim install 'mmdet>=3.0.0'
110+
mim install 'mmdet>=3.0.0rc5, <3.3.0'
111111
```
112112

113113
**Step 2.** Install mmdetection3d forked repository
@@ -144,7 +144,7 @@ python tools/create_data.py nuscenes --root-path ./data/nuscenes --out-dir ./dat
144144
#### Prepare the config file
145145

146146
The configuration file that illustrates how to train the CenterPoint model with the NuScenes dataset is
147-
located at mmdetection3d/configs/centerpoint/centerpoint_custom.py. This configuration file is a derived version of the
147+
located at mmdetection3d/projects/AutowareCenterPoint/configs. This configuration file is a derived version of the
148148
centerpoint_pillar02_second_secfpn_head-circlenms_8xb4-cyclic-20e_nus-3d.py configuration file from mmdetection3D.
149149
In this custom configuration, the **use_voxel_center_z parameter** is set to **False** to deactivate the z coordinate of the voxel center,
150150
aligning with the original paper's specifications and making the model compatible with Autoware. Additionally, the filter size is set as **[32, 32]**.
@@ -155,7 +155,7 @@ This includes adjustments related to preprocessing operations, training, testing
155155
#### Start training
156156

157157
```bash
158-
python tools/train.py configs/centerpoint/centerpoint_custom.py --work-dir ./work_dirs/centerpoint_custom
158+
python tools/train.py projects/AutowareCenterPoint/configs/centerpoint_custom.py --work-dir ./work_dirs/centerpoint_custom
159159
```
160160

161161
#### Evaluation of the trained model
@@ -188,7 +188,7 @@ python tools/create_data.py Tier4Dataset --root-path data/sample_dataset/ --out-
188188
Run evaluation
189189

190190
```bash
191-
python tools/test.py ./configs/centerpoint/test-centerpoint.py /PATH/OF/THE/CHECKPOINT --task lidar_det
191+
python tools/test.py projects/AutowareCenterPoint/configs/centerpoint_custom_test.py /PATH/OF/THE/CHECKPOINT --task lidar_det
192192
```
193193

194194
Evaluation results could be relatively low because of the e to variations in sensor modalities between the sample dataset
@@ -203,10 +203,10 @@ the base link location of the vehicle.
203203
The lidar_centerpoint implementation requires two ONNX models as input the voxel encoder and the backbone-neck-head of the CenterPoint model, other aspects of the network,
204204
such as preprocessing operations, are implemented externally. Under the fork of the mmdetection3d repository,
205205
we have included a script that converts the CenterPoint model to Autoware compatible ONNX format.
206-
You can find it in `mmdetection3d/tools/centerpoint_onnx_converter.py` file.
206+
You can find it in `mmdetection3d/projects/AutowareCenterPoint` file.
207207

208208
```bash
209-
python tools/centerpoint_onnx_converter.py --cfg configs/centerpoint/centerpoint_custom.py --ckpt work_dirs/centerpoint_custom/YOUR_BEST_MODEL.pth -work-dir ./work_dirs/onnx_models
209+
python projects/AutowareCenterPoint/centerpoint_onnx_converter.py --cfg projects/AutowareCenterPoint/configs/centerpoint_custom.py --ckpt work_dirs/centerpoint_custom/YOUR_BEST_MODEL.pth --work-dir ./work_dirs/onnx_models
210210
```
211211

212212
#### Create the config file for the custom model

0 commit comments

Comments
 (0)