yoxu515

Hi, lifa, I am training it from scratch on S3DIS and got approximately the same results as yours. Specifically, 30 epochs now (still training), best mIoU 58.388, best eval acc...

Hi, @JerryIndus @lqzhao , I didn't pay much effort to fix this repo. I tried another repo of RandLANet with pytorch, and you can refer to this [repo](https://github.com/qiqihaer/RandLA-Net-pytorch). But as...

same problem!

Hi, can you describe what error did you get?

Hello, sorry for the error. This bug exits when the "pytorch correlation" is not correctly installed, similar in issue #45. I recommend you to install it and the code also...

Hi, please check [here](https://github.com/yoxu515/aot-benchmark#demo---panoptic-propagation) in README, which allows you to run models on custom datasets. The results are saved as videos and masks. - We haven't release the code of...

Hi, MS-AOT currently is not integrated in this repository. The code in issue #32 is only for vot datasets and different from the framework in this repo. We will integrate...

Hi, thank you for pointing out the issue. Could you please give an image for example? where does the 0.5px/1px misaligment bias happen?

Hi, in the demo.ipynb, by default the output_video has the same format as the input video, i.e. 'output_video': './assets/pl_seg.mp4', since the default fourcc is obtained from the source video. If...

> > Hi, in the demo.ipynb, by default the output_video has the same format as the input video, i.e. 'output_video': './assets/pl_seg.mp4', since the default fourcc is obtained from the source...