kazunaritakeichi

Results 8 comments of kazunaritakeichi

Hi @lunzueta Thank you! Ok, I'll also try and tell you!

Hi @lunzueta. One way against overfittin may be tuning dropout parameters. http://stats.stackexchange.com/questions/109976/in-convolutional-neural-networks-how-to-prevent-the-overfitting

@lunzueta σ parameter is set to 1.0 for FLIC and 2.0 for LSP in the paper and lr parameter is set to 0.0005 for both datasets, isn't it? I don't...

@lunzueta This is log.png (cropping is 0). test loss is increasing... ![log](https://cloud.githubusercontent.com/assets/13772320/16460672/793abca6-3e63-11e6-96c8-27313bc7095e.png)

I tried newer version(shell/train_lsp.sh). Below is the result. ![log](https://cloud.githubusercontent.com/assets/13772320/16536159/5e34ef9a-4029-11e6-9eea-469573a92819.png)

Thank you so much @mitmul ! I'll learn the paper and the code so that I can contribute.

I tried with FLIC dataset. I got the similar result with MPII @lunzueta.

@lumaku Thank you for your reply. I specify the version like `ctc-segmentation = "1.7.4"` in `pyproject.toml`. So, I think the package is installed from PyPl. `poetry.lock` is like below. ```...