Yuecong Min
Yuecong Min
Yes, it seems there is a typo here, and the corresponding function should be uvd2xyz_nvidia() , thanks for your correction! During our experiments, we didn't notice significant performance difference between...
> Hi, can you share the code of training NvGesture dataset? I found that only replacing the dataset in the pointlstm.yaml file does not achieve the performance in your paper....
Hi, @ceb-jjgf , The train_gesture.txt is the input list of SHREC'17 dataset, you can download this datase from the website: [SHREC](http://www-rech.telecom-lille.fr/shrec2017-hand/). Each line follow the following pattern: id_gesture id_finger id_subject...
@hamza-hadda As we mentioned in that paper, the front camera of HUAWEI mate 20 and 30 is accompanied by a secondary sensor for depth sensing.
You can check the return value of r.split(npy_path), npy_path and label are expected to obtain from it.
It seems that "nvidia_dataloader.py" has 32 lines.
OK, I got it. It is in the [nvidia_dataloader.py](https://github.com/ycmin95/pointlstm-gesture-recognition-pytorch/blob/4f65853aa6e57a30b4620830f021dfcf6ab442e5/experiments/nvidia_dataloader.py#L106) file, which is used for debugging in the previous version, you can simply change the BaseFeeder to NvidiaFeeder. But it should...
Can you provide more details about the experiments (e.g., the experimental log)? What kind of performance can you achieve?
Hi, it has been a long time from this experiment, and I remember that we select this dataset to show the generalization and do not perform ablation study on this...
Sorry for late reply, you can visualize the point cloud sequence at different stages, which should be likely to Figure 4. If you will, I can send the source data...