[Doc] for Visualization feature map using wandb backend in dev-1.x #2557
Conversation
|
It looks good to me.
|
|
Looks good to me except changing |
jinxianwei
requested review from
MeowZheng and
MengzhangLI
and removed request for
MeowZheng
| from mmseg.apis import inference_model, init_model | ||
| from mmseg.structures import SegDataSample | ||
| from mmseg.utils import register_all_modules | ||
| from mmseg.visualization import SegLocalVisualizer |
There was a problem hiding this comment.
hi,could you tell me how to install these package from 51 to 54,thanks!
There was a problem hiding this comment.
hi,could you tell me how to install these package from 51 to 54,thanks!
These path would be automatically searched in your mmseg codebase (Note that it is dev-1.x branch rather than master branch). You just need to follow the guidance in this README, create a python file feature_map_visual.py and put it in this place ./mmsegmentation/feature_map_visual.py.
There was a problem hiding this comment.
dev-1.x docs link
https://mmsegmentation.readthedocs.io/en/dev-1.x/get_started.html
There was a problem hiding this comment.
hi,I've try reinstall the whole envriment as this doc describe,now my mmcv version is 2.0.0rc4,and mmengine0.5.0,mmsegmentation1.0.0rc5.but 'vidualize_feature_map.py' line 106,'inference' line 119, my 'data_'has no atrribute '['inputs']'(use 'loadimagefromfile').also,in 'mmcv.transform.loadimagefromfile',there is no '['inputs']'. so can you tell me your mmcv version?
There was a problem hiding this comment.
hi,I've try reinstall the whole envriment as this doc describe,now my mmcv version is 2.0.0rc4,and mmengine0.5.0,mmsegmentation1.0.0rc5.but 'vidualize_feature_map.py' line 106,'inference' line 119, my 'data_'has no atrribute '['inputs']'(use 'loadimagefromfile').also,in 'mmcv.transform.loadimagefromfile',there is no '['inputs']'. so can you tell me your mmcv version?
I also use 2.0.0rc4 mmcv
There was a problem hiding this comment.
I didn't fix it,but still thank you.
There was a problem hiding this comment.
I I followed your method to visualize the heat map, but it does not show the heat map because recorder.data_buffer=0. What can I do to fix it?
There was a problem hiding this comment.
you can verify the sub module name of "source" in your model by debug.
There was a problem hiding this comment.
you can verify the sub module name of "source" in your model by debug.
In a semantic partitioned heat map, usually one heat map is drawn for one class. I have used your method to draw a heat map that contains all classes, and the function to draw the heat map in mmengine does not seem to have a parameter to select a class.
|
yangyang-luo702，您好！您在MMSeg项目中给我们提的PR非常重要，感谢您付出私人时间帮助改进开源项目，相信很多开发者会从你的PR中受益。 |
…pen-mmlab#2557) ## Motivation Docs for Visualization featusre map using wandb backend. ## Modification Add a new markdown file and result demo of wandb. --------- Co-authored-by: MeowZheng <meowzheng@outlook.com>
Thanks for your contribution and we appreciate it a lot. The following instructions would make your pull request more healthy and more easily get feedback. If you do not understand some items, don't worry, just make the pull request and seek help from maintainers.
Motivation
Docs for Visualization featusre map using wandb backend.
Modification
Add a new markdown file and result demo of wandb.