Eric Jau

Hi @sanjayr1980 , Is it all the error you got? I didn't really see what happened here. Maybe you can paste more complete logs?

Hi @YanShuo1992 , Thanks for your message. To generate keypoints for training, I suggest directly using the pre-trained network for homography adaptation (https://github.com/eric-yyjau/pytorch-superpoint/tree/master/logs/magicpoint_synth20/checkpoints). For the parameters, you may refer to...

Hi @YanShuo1992 , Yeah I remember I saw this issue before but I don't know what caused this. I think some pytorch version will have this issue. If you know...

Hi @35p32, Thanks for your question. Sorry I don't know why it happened. Maybe it is because of the pytorch version. If you have a way to prevent this from...

Hi @saunair , Thanks a lot. Sure I can review them. Could you give more description on the code change (Maybe a commit with all of your change) and the...

Hi @jcyhcs, (Thanks but I'm not a professor though). This tutorial may help (https://pytorch.org/tutorials/advanced/super_resolution_with_onnxruntime.html). This repo does not cover exporting to onnx but I imagine the general pytorch method works.

Hi @JuntingLee, Thank you for your question. I'm not sure why this happened. Can you check if the points are loaded correctly? https://github.com/eric-yyjau/pytorch-superpoint/blob/4ff74df8fa3c10ce9eb9fdc561f787d9e8bc9691/datasets/Coco.py#L290 There may be some weird thing happening...

Hi @rishabhramteke , I think the keypoints can be obtained from the output of this line. https://github.com/eric-yyjau/pytorch-superpoint/blob/master/Train_model_frontend.py#L774 Hope this helps. Thanks.

Hi @SuanNaiShuiGuoLao , Thank you for your question. I'm not sure what causes the difference. In the paper, it says "Repeatability is computed at 240 x 320 resolution with 300...

Looks like the dataset path is not set correctly? So there's no input for the dataloader. Thanks.