mmdetection3d在win10下实现点云目标检测

主要用官网demo, 查看检测的效果(准确度)

1. 样列

单模态样列

python demo/pcd_demo.py demo/data/kitti/kitti_000008.bin configs/pointpillars/hv_pointpillars_secfpn_6x8_160e_kitti-3d-car.py checkpoints/hv_pointpillars_secfpn_6x8_160e_kitti-3d-car_20220331_134606-d42d15ed.pth --device cuda:0 --score-thr 0.7 --out-dir output/demo_test --show

多模态样列

python demo/multi_modality_demo.py demo/data/kitti/kitti_000008.bin demo/data/kitti/kitti_000008.png demo/data/kitti/kitti_000008_infos.pkl configs/mvxnet/dv_mvx-fpn_second_secfpn_adamw_2x8_80e_kitti-3d-3class.py checkpoints/dv_mvx-fpn_second_secfpn_adamw_2x8_80e_kitti-3d-3class_20210831_060805-83442923.pth --out-dir output/demo_test --show

3D分割

python demo/pc_seg_demo.py demo/data/scannet/scene0000_00.bin configs/pointnet2/pointnet2_ssg_16x2_cosine_200e_scannet_seg-3d-20class.py checkpoints/pointnet2_ssg_16x2_cosine_200e_scannet_seg-3d-20class_20210514_143644-ee73704a.pth

2. 已有模型在标准数据集上推理和训练

本电脑只有一个gpu, 单块显卡测试数据集: ​​

python tools/test.py ${CONFIG_FILE} ${CHECKPOINT_FILE} [--out ${RESULT_FILE}] [--eval ${EVAL_METRICS}] [--show] [--show-dir ${SHOW_DIR}]

(1) 保存模型，可视化结果
python tools/test.py configs/pointpillars/hv_pointpillars_secfpn_6x8_160e_kitti-3d-car.py checkpoints/hv_pointpillars_secfpn_6x8_160e_kitti-3d-car_20220331_134606-d42d15ed.pth --show --show-dir output/demo_test/show_results

(2) 不保存测试结果，计算mAP
python tools/test.py configs/pointpillars/hv_pointpillars_secfpn_6x8_160e_kitti-3d-car.py checkpoints/hv_pointpillars_secfpn_6x8_160e_kitti-3d-car_20220331_134606-d42d15ed.pth --eval mAP

(3) 保存模型，可视化结果，可视化真实标签，计算mAP
python tools/test.py configs/pointpillars/hv_pointpillars_secfpn_6x8_160e_kitti-3d-car.py checkpoints/hv_pointpillars_secfpn_6x8_160e_kitti-3d-car_20220331_134606-d42d15ed.pth --eval mAP --eval-options 'show=True' 'out_dir=./data/test/show_results'

问题

ETA:D:\PythonVirtualEnv\envs\mmdet3d\lib\site-packages\torch\functional.py:568: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at C:\actions-runner\_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\TensorShape.cpp:2228.)

解决方案在对应D:\PythonVirtualEnv\envs\mmdet3d\lib\site-packages\torch\functional.py文件的568行进行修改。

目前这行代码依然会报错。后续将继续完善

python tools/test.py configs/votenet/votenet_8x8_scannet-3d-18class.py checkpoints/votenet/votenet_8x8_scannet-3d-18class_20200620_230238-2cea9c3a.pth --eval mAP --eval-options 'show=True' 'out_dir=./data/scannet/show_results'