Inference demo

[Lmorring]

Excellent work! May I ask if you have plans to provide a demo script for inference?

[YilmazKadir]

Can you describe what you mean in more detail such that we can help you? Do you want to run inference on a self-collected point cloud?

[Lmorring]

yes, but I've solved it. Thank you for your answer.

[2003416]

我能问一下你是怎么解决的吗，不知道为啥我用预训练的权重跑出了在semantickitti上的结果只有0.5876，不知道是哪出问题了。我用的测试代码是：python tools/test.py --config-file configs/semantic_kitti/semseg-pt-v3m1-0-image.py --options save_path=exp/semantic_kitti/semseg-pt-v3m1-0-image weight=pertrain/model/model_best.pth

[YilmazKadir]

You need to download KITTI images and link it to data/semantic_kitti_color. You can use this link to download the images: https://s3.eu-central-1.amazonaws.com/avg-kitti/data_odometry_color.zip

[2003416]

You need to download KITTI images and link it to data/semantic_kitti_images. You can use this link to download the images: https://s3.eu-central-1.amazonaws.com/avg-kitti/data_odometry_color.zip

Thank you for your reply. This shows the contents of my data folder. I have already placed the image data inside the corresponding directories before running the test code.
data/
├── semantic_kitti/
│ └── dataset/
│ └── sequences/
│ ├── 00/
│ ├── 01/
│ ├── 02/
│ ├── 03/
│ ├── 04/
│ ├── 05/
│ ├── 06/
│ ├── 07/
│ ├── 08/
│ │ ├── image_2/
│ │ ├── image_3/
│ │ ├── labels/
│ │ ├── velodyne/
│ │ ├── calib.txt
│ │ ├── poses.txt
│ │ └── times.txt
│ ├── 09/
│ └── 10/
└── semantic_kitti_color/
└── dataset/
└── sequences/
├── image_2/
├── image_3/
└── .gitkeep

[vvonjoon]

Can you describe what you mean in more detail such that we can help you? Do you want to run inference on a self-collected point cloud?

Any plans for an inference(demo) code for self-collected images and point clouds?

[YilmazKadir]

Sure. I will work on it and get back you soon.