Merge pull request #1093 from Cedarch/master

update readme of MPPNet

README.md

@@ -23,7 +23,7 @@ It is also the official code release of [`[PointRCNN]`](https://arxiv.org/abs/18
 
 ## Changelog
 [2022-09-02] **NEW:** Update `OpenPCDet` to v0.6.0:
-* Official code release of [MPPNet](https://arxiv.org/abs/2205.05979) for temporal 3D object detection, which supports long-term multi-frame 3D object detection and ranks 1st place on [3D detection learderboard](https://waymo.com/open/challenges/2020/3d-detection) of Waymo Open Dataset on Sept. 2th, 2022 (see the [guideline](docs/guidelines_of_approaches/mppnet.md) on how to train/test with MPPNet).
+* Official code release of [MPPNet](https://arxiv.org/abs/2205.05979) for temporal 3D object detection, which supports long-term multi-frame 3D object detection and ranks 1st place on [3D detection learderboard](https://waymo.com/open/challenges/2020/3d-detection) of Waymo Open Dataset on Sept. 2th, 2022. For validation dataset, MPPNet achieves 74.96%, 75.06% and 74.52% for vehicle, pedestrian and cyclist classes in terms of mAPH@Level_2. (see the [guideline](docs/guidelines_of_approaches/mppnet.md) on how to train/test with MPPNet).
 * Support multi-frame training/testing on Waymo Open Dataset (see the [change log](docs/changelog.md) for more details on how to process data).
 * Support to save changing training details (e.g., loss, iter, epoch) to file (previous tqdm progress bar is still supported by using `--use_tqdm_to_record`). Please use `pip install gpustat` if you also want to log the GPU related information.
 * Support to save latest model every 5 mintues, so you can restore the model training from latest status instead of previous epoch.   

docs/guidelines_of_approaches/mppnet.md

+## NOTE
+**If you want to quickly develop your own model based on MPPNet, our recommended setting is to use mppnet_4frames.yaml, disable `USE_ROI_AUG` and `USE_TRAJ_AUG` flags in yaml and train 3 epoch. A reference time cost for this setting is about 5 hours, using 8 A100 GPUs.  After finishing your development, you can get stable gains when using mppnet_16frames.yaml, enabling `USE_ROI_AUG` and `USE_TRAJ_AUG` flags and training 6 epoch.**
+
 ## Installation
 
 Please refer to [INSTALL.md](docs/INSTALL.md) for the installation of `OpenPCDet`.
 
-
-##Data Preparation
+## Data Preparation
 Please refer to [GETTING_STARTED.md](docs/GETTING_STARTED.md) to process the Waymo Open Dataset.
 
-##Training
+## Training
 
 1.  Train the RPN model  for MPPNet (centerpoint_4frames is employed in the paper)
 ```shell
@@ -35,11 +37,12 @@ bash scripts/dist_train.sh ${NUM_GPUS} --cfg_file cfgs/waymo_models/mppnet_4fram
 --set DATA_CONFIG.ROI_BOXES_PATH.train  ../output/waymo_models/centerpoint_4frames/default/eval/epoch_36/train/default/result.pkl \
       DATA_CONFIG.ROI_BOXES_PATH.test  ../output/waymo_models/centerpoint_4frames/default/eval/epoch_36/val/default/result.pkl
 ```
-When using 16-frame, we can just change the `cfg_file` to mpppnet_16frames.yaml and the ` DATA_CONFIG.ROI_BOXES_PATH`  is same with 4-frame.
-We can also save the paths of train and val results to ROI_BOXES_PATH in mppnet_4frames.yaml and mppnet_16frames.yaml to avoid using the `set` flag.
-For each GPU, BATCH_SIZE should be at least equal to 2.  When using 16-frame, the reference GPU memory consumption is 29G with BATCH_SIZE=2.
+When using 16-frame, we can just change the `cfg_file` to mpppnet_16frames.yaml and the `DATA_CONFIG.ROI_BOXES_PATH` is same with 4-frame.\
+We can also save the paths of train and val results to ROI_BOXES_PATH in mppnet_4frames.yaml and mppnet_16frames.yaml to avoid using the `set` flag.\
+For each GPU, BATCH_SIZE should be at least equal to 2.  When using 16-frame, the reference GPU memory consumption is 29G with BATCH_SIZE=2.\
+**Note**: Disable the `USE_ROI_AUG` and `USE_TRAJ_AUG` flag in config yaml can double the training speed with a performance loss of about 0.4%. 
 
-##Evaluation
+## Evaluation
 * Test with a pretrained model:
 ```shell
 # Single GPU
@@ -59,3 +62,12 @@ python test.py --cfg_file cfgs/waymo_models/mppnet_e2e_memorybank_inference.yaml
 --pretrained_model  ../output/waymo_models/centerpoint_4frames/default/ckpt/checkpoint_epoch_36.pth
 ```
 The default parameters in mppnet_e2e_memorybank_inference.yaml is for 4-frame and just change them to the setting in mppnet_16frames.yaml when using 16-frame.
+
+## Performance
+|    Model          | Vec_L1 | Vec_L2 | Ped_L1 | Ped_L2 | Cyc_L1 | Cyc_L2 |  
+|:---------------------------------------------:|:----------:|:-------:|:-------:|:-------:|:-------:|:-------:|
+|  [centerpoint_4frames](../../tools/cfgs/waymo_models/centerpoint_4frames.yaml) | 76.71/76.17 | 69.13/68.63 | 78.88/75.55 | 71.73/68.61 | 73.73/72.96 | 71.63/70.89 |
+|  [mppnet_4frames](../../tools/cfgs/waymo_models/mppnet_4frames.yaml) | 81.54/81.06 | 74.07/73.61 | 84.56/81.94 | 77.20/74.67 | 77.15/76.50 | 75.01/74.38 |
+| [mppnet_16frames](../../tools/cfgs/waymo_models/mppnet_16frames.yaml) | 82.74/82.28 | 75.41/74.96 | 84.69/82.25 | 77.43/75.06 | 77.28/76.66 | 75.13/74.52 |
+
+The reported performance of MPPNet is trained with 6 epoch with  `USE_ROI_AUG`  and  `USE_TRAJ_AUG`  flags enabled.
\ No newline at end of file