Open shinianzhihou opened 2 years ago
hello bro, i am facing the same problem with you, can we solve it together.
Hi, thanks for your issue and feedback! @zzwj66 @shinianzhihou
Could you please some paper, dataset or github repo for us? Because we do not know too much about this field.
I mean we could start from review and roadmap, thank you so much.
Best,
paper-> https://arxiv.org/pdf/2011.06961.pdf (from arxiv) code-> modified based on mmseg, it mainly rewrite the pipeline and dataset while some bugs(cannot val when training) in it.
@MengzhangLI
paper with code ->https://paperswithcode.com/paper/deep-multimodal-fusion-by-channel-exchanging i want to deal with NYUDV2 dataset
OK, I will record your provided infomation and have a look. Thank you.
I noticed that MultiImageMixDataset have been released, is it related to this issue? @MengzhangLI
Has this issue been solved?
Has this issue been solved?
Not yet. I have rewritten dataset and model for my task, ChangeDetection.
Hi~, has this issue been resolved? Will mmseg support such multi-input single-output models? @MengzhangLI
Has anyone studied mixed input of RGB and RGB-T?
Hi, I want to train multispectral and thermal. Has anyone studied, solved?
Hi~, has this issue been resolved? Will mmseg support such multi-input single-output models? @MengzhangLI
Has this issue been solved?
hey all, has this issue been solved?
Has this issue been solved?
Not yet. I have rewritten dataset and model for my task, ChangeDetection.
i am trying to solve this(multimodal input) now, but i met lots of questions. Could we discuss this together? Maybe you can leave your wechat number?
Hi~, has this issue been resolved? Will mmseg support such multi-input single-output models? @MengzhangLI
Could we discuss it together? I am trying to do this now
How to make Multi-Input-Single-Output segmentation?
Could you please provide some suggestions about how to handle multi-input such as depth and rgb, while to the maximum keeping the mmseg structure?
Thanks!