Li Jiang

icon company The Chinese University of Hong Kong

Results 4 comments of Li Jiang

Scannet v2

Hi, You don't need to download the whole dataset. You could just download `[scene_id]_vh_clean_2.ply`, `[scene_id]_vh_clean_2.labels.ply`, `[scene_id]_vh_clean_2.0.010000.segs.json` and `[scene_id].aggregation.json` files.

Can I use python3.6 instead of python3.7?

Hi, I don't test with python3.6, but I think it should also be applicable.

Training with own data

Hi, The data loaded from the file for a scene [[here]](https://github.com/Jia-Research-Lab/PointGroup/blob/dea7ab1fbe6c2a519a251beac595085295439f77/data/scannetv2_inst.py#L176) is of the form: * xyz: (N, 3), point coordinates * rgb: (N, 3), rgb values normalized to (-1,...

grad_output.continuous()

Hi, You can directly use SparseConvNet instead of spconv. No other change is needed.