Xudong Cai

> Got very unsatisfactory on these kind of images. Am I doing something wrong? > >  Hi，I want to ask you that where can I get the train dataset?...

> I have followed the steps to run demo but I get inaccurate results. The boxes of heads are seen to be offset from their original coordinate(see the results). >...

你好，有没有更新后的源码可以放出来一下，方便大家共同学习和维护

> Thanks for your great work, the recall@1% on Oxford is higher than that in the paper, which is trained by the default config in your repository. Is it normal?...

> Lastly you need to typecast to dict in line 47 in LoFTR/src/datasets/megadepth.py: > > ``` > self.intrinsics = dict(np.load(intrinsic_path)) > ``` > > Hope everything works! Thanks for your...

I'm training `outdoor_ds` with the default setting(image size 640), I use 4 3090 GPUs too. Also, i use the origin megadepth data to train as the undistorted image is not...

After 30 epoches training, I reproduced the test on megadepth and got: 'auc@10': 0.6676607412455137, 'auc@20': 0.7952598445093988, 'auc@5': 0.4983204021567033, 'prec@5e-04': 0.9549532078302655 3 points lower than the reported accuracy.

>  @Master-cai Hello，This is the training result I obtained when I set 'TRAIN_IMG_SIZE' to 640. Your training result is much better than mine. Have you tried setting 'TRAIN_IMG_SIZE' to...

@Mysophobias I didn't process the megadepth via D2-net, and your ckpts seems similar to mine. I have no idea about your bad test results. I just used the default `reproduce_test\outdoor_ds.sh`...

@Mysophobias 3090 can train it with physical bs=1. I use accumulate grad=2 to make the bs=1 * 2 * 4=8, which is suggested by the author. i have trained it...