Guolei Sun

icon indicating a website link https://guoleisun.github.io

icon company ETH Zurich icon location Switzerland icon location Postdoc at Computer Vision Lab, ETH Zurich

Results 14 comments of Guolei Sun

Dose the test reuslts on several images rather than whole videos represent the performance of video semantic segmentation methods?

Hi, thanks for your interest. For [VSPW dataset](https://yu-wu.net/pdf/CVPR21-vspw.pdf), the test performance is on the **_whole_** videos, rather than several images. For cityscapes, it is true that the test performance is...

Accuracy of the training

Hi, I also meet the same problem, getting error of 18.3% using ResNeXt-29, 8x64 model.

Code for video visualization

Thanks for your interest. When you use our model for evaluation, you could add "--format-only --eval None" to your command. The code is largely based on [MMseg](https://github.com/open-mmlab/mmsegmentation) and the usage...

Resnet50 CUB 2011 performance

@jeong-tae I have the same problem reproducing the results of resnet50 (84.5). I also got less than 80% using resent50. @Lzc6996 Can you be more specific when you refer us...