Question about New? Model & OFD trick

[hirokic5]

Thank you for your great research & repository, and congrats on MODNet acceptance for AAAI 2022 !!! ( I firts read your MODNet paper about 1 year ago, and I have been looking forward to training code release.)

Now, I have some questions.

1. Newer Model You said in README.md ...

A Single model! Only 7M! Process 2K resolution image with a Fast speed on common PCs or Mobiles! Beter than research demos!

By the way, the pretrained model from this link seems to be about 25M. So, did you train newer model in 2022 ? and do you have any plans for release that model ?

2. OFD trick In the paper, OFD trick is used for inference. Do you have any plans for release the OFD trick ?

Thanks.

[ZHKKKe]

Hi, thanks for your attention.

1. The 7M model will not be released. It is trained on a larger dataset and not used for research purpose.
2. The implement of OFD is not complex. I may release it soon (I am busy recently, and this may be delayed for a while).

[zhanghongyong123456]

Hi, thanks for your attention.

1. The 7M model will not be released. It is trained on a larger dataset and not used for research purpose.
2. The implement of OFD is not complex. I may release it soon (I am busy recently, and this may be delayed for a while).

您好，请问关于OFD代码，是否会发布