search

SimonVandenhende / Multi-Task-Learning-PyTorch

PyTorch implementation of multi-task learning architectures, incl. MTI-Net (ECCV2020).
Other
780 stars 113 forks source link

ResNet50 and PAD-net #13

Closed rongmq8802 closed 3 years ago

rongmq8802 commented 3 years ago

First of all, thank you for opening up the code. I am learning multi-task learning recently. And from your paper "Multi-Task Learning for Dense Prediction Tasks: A Survey", it can be concluded that on NYDU-v2 dataset, the PAD-Net based on ResNet-50 seems to get the best results, but your program does not support this situation. May I ask if you can provide one or whether the current code can achieve this function.

SimonVandenhende commented 3 years ago

Hello, thanks for the interest.

Indeed, the current code does not include PAD-Net using a ResNet-50 backbone. I took the number from the original paper, as the overall settings are very similar. From my personal experience, I found it difficult to reproduce the exact numbers from the PAD-Net paper. It might be better to talk to the authors directly about this. Note that I did spend some time tuning the HRNet-based version of PAD-Net, so that implementation should be fine I believe.

rongmq8802 commented 3 years ago

Ok, I got it. Thanks a lot！