Our one paper On Symmetric Losses for Learning from Corrupted Labels has accepted to ICML 2019.