Frost

`conda install tensorflow-gpu`

Upgrade to PyTorch 1.3 can solve this problem.

@Yxt1212 I met the same issue. It is solved by `_CC.SOLVER.WEIGHT_DECAY_BIAS = 0.0 # None` in `defrcn/config/defaults.py`

@foralliance Better than me. I also used fake annotations but got 0.298 mAP online. Are there any good ways to test VOC2012?

@XuanheLiu This is not possible

@DEepLiker When you run `main-run-twoStream.py`, do you change any parameters? I got only 71.55% for joint training. Now I am testing different hyper-parameters with the hope that the score can...

Guys sorry for the delay! Here is the update. The G-TAD+PGCN experiment was implemented on another machine, but I have a chance to find the script that I used for...

Hi @RonanDou The "detection_results.txt" file is the txt format of the json file after post-processing. You may need to change the line here. https://github.com/frostinassiky/gtad/blob/6deb5b1bc6883b48bd22e0cc593069643c953e3d/gtad_postprocess.py#L154-L156

Hi @RonanDou As I remember, I changed the NMS threshold and weights to combine RBG and optical flow results. Since you find the P-GCN result is worse than G-TAD, what...

Hi @Oliver0804 Please upgrade your CUDA to 10.0+ and try to install it again.