ZhengdiYu
commented
1 year ago First of all, thank you for your awesome work! Also, thank you for kindly sharing your code and model weights to facilitate future research.
I'm currently working on a research project (which I'm targeting to submit in May) considering your work as one of the baseline methods. However, I have been struggling with reproducing your results on InterHand2.6M dataset (as reported in the paper).
In the other thread, you mentioned that you used other extra in-the-wild datasets during training for the model weights currently shared in this repo. So, I was wondering if you used different model weights (trained only on InterHand2.6M) when obtaining the InterHand2.6M results in the paper. If so, would you kindly share the weights for InterHand2.6M dataset as well?
Thank you so much! 👍
Thank you so much for your interest in our work! :)
Yes, the current pre-trained weights are just used for the in-the-wild demo.
Please rest assured, the training/eval code and weights will also be released together soon, after we restructure the code into this simple & clean version, hopefully before May!
jyunlee
First of all, thank you for your awesome work! Also, thank you for kindly sharing your code and model weights to facilitate future research.
I'm currently working on a research project (which I'm targeting to submit in May) considering your work as one of the baseline methods. However, I have been struggling with reproducing your results on InterHand2.6M dataset (as reported in the paper).
In the other thread, you mentioned that you used other extra in-the-wild datasets during training for the model weights currently shared in this repo. So, I was wondering if you used different model weights (trained only on InterHand2.6M) when obtaining the InterHand2.6M results in the paper. If so, would you kindly share the weights for InterHand2.6M dataset as well?
Thank you so much! 👍