search

datvuthanh / HybridNets

HybridNets: End-to-End Perception Network
MIT License
594 stars 121 forks source link

Drivable and Lane Type training #21

Closed abhigoku10 closed 2 years ago

abhigoku10 commented 2 years ago

@datvuthanh thanks for sharing the code based i have following queries

  1. since bdd100k has different types of lanes eg double yellow, single white lane , dashed, can we use current source code to train for different lane types? is so what are the modifications need to be made int he code based
  2. Can we similar train the current source code with the driveable area and alternate drivable area labels ? if so what are the changes to be made

Please share your thoughts Thanks in advance

datvuthanh commented 2 years ago

Yes, I don't think it is problem. On segmentation task, the imbalanced classes is a challenge and we solved this problem by combining between Focal Loss and Tversky Loss. You can try it and adjust parameters a, b: Loss = a * Focal Loss + b * Tversky Loss

xoiga123 commented 2 years ago

Yes, I don't think it is problem. On segmentation task, the imbalanced classes is a challenge and we solved this problem by combining between Focal Loss and Tversky Loss. You can try it and adjust parameters a, b: Loss = a * Focal Loss + b * Tversky Loss

I think he's correct 👍 Linking to #20 for easy tracking.