bearcatt
bearcatt
@tianzhi0549 I've tried sharing the center-ness head with the reg head rather than the cls head. This simple modification leads to 0.5 mAP gain for FCOS-R101. You can try this.
@tianzhi0549 that's great! You can also refer to the code [here](https://github.com/HRNet/HRNet-FCOS/blob/master/maskrcnn_benchmark/modeling/rpn/fcos/fcos.py#L88).
I saw the commit message "update tensorflow to version 1.4", does it mean that we can use tensorflow_fold in tensorflow 1.4?
> In case I try my own implementation, adding length-level embedding to an input word embedding will suffice? Yes, that's it. > release the code/model for Length-Controllable VLP It's ok...
Hi, thanks for your interest! The length-controllable version of AoANet can be implemented by simply adding a length-level embedding to the input word embedding, and the other settings directly follow...
@RubickH @dXDb https://github.com/bearcatt/Length-Controllable-AoANet
Hi @yulunzhang, have you update the code?
@lt1103725556 in Table1, the models trained with 100 proposals are evaluated with 100 proposals, which is the correct setting. However, for models that are marked with "*" (i.e., trained with...
Hi @CR-Gjx , in Table 3 of your paper, how do you obtain the results of SeqGAN and RankGAN? I can't find any result about image captioning in the paper...