This paper presents a new mechanism to facilitate the training of mask
transformers for efficient panoptic segmentation, democratizing its deployment.
We observe that due to its high complexity, the training objective of panoptic
segmentation will inevitably lead to much higher false positive penalization.
Such unbalanced loss makes the training process of the end-to-end
mask-transformer based architectures difficult, especially for efficient
models. In this paper, we present ReMaX that adds relaxation to mask
predictions and class predictions during training for panoptic segmentation. We
demonstrate that via these simple relaxation techniques during training, our
model can be consistently improved by a clear margin \textbf{without} any extra
computational cost on inference. By combining our method with efficient
backbones like MobileNetV3-Small, our method achieves new state-of-the-art
results for efficient panoptic segmentation on COCO, ADE20K and Cityscapes.
Code and pre-trained checkpoints will be available at
https://github.com/google-research/deeplab2.

本文提出一种 ReMaX 的新机制，通过在学习时添加对分割蒙版和类别蒙版的松弛约束，可以实现更高效的全景分割，使其更易于部署，同时可以与 MobileNetV3-Small 等高效 backbones 结合使用，从而在 COCO、ADE20K 和 Cityscapes 等数据集上取得最新的最新的最好成果。