Ke Young

Is there any difference with the official repo?

2

Hi , thanks very much for your efforts. I noticed that this is not the official repo. Could you please tell me what's the difference between them, such as implementation...

Any plan for Swin Transformer?

1

### Describe the feature Hi there! Is there any plan for Swin Transformer?

How about the performance？

6

Hi， Have you test the model with standard VOC evaluation protocol?

Hi Yunhang, Could you give me an approximate time that you will release the codes? Thanks very much.

About the performance.

3

It is weird that the mAP I got with your codes is about 37~38. I use the default setting. I also can not reproduce the results of https://github.com/luiszeni/Boosted-OICR.

prediction for a single new image

1

Thank you very much for contributing such a good image annotation code, how can I make a prediction for a single new image?

Does this repo support pre-computed proposals?

1

Hi, Does this repo support pre-computed proposals? BTW， have you tested on Nvidia RTX3090 GPU?

Does VIDEO_SEGMENTATION_DATA data format still exist for v1.1？

1

Does VIDEO_SEGMENTATION_DATA data format still exist for v1.1？

For visualization， the energy value of every time step is needed. I read the codes, it seems I need to change somewhere in the recurrentshop lib.

Hi, Thanks for sharing this code repo. I am doubtful how are the thermal and depth backbone initialized. Randomly or with imagenet-pretrained weights? Thanks!