DABnet:Detaching and Boosting: Dual Engine for Scale-Invariant Self-Supervised Monocular Depth Estimation
This is the code for monocular self-supervised depth estimation model using the method described in
**Detaching and Boosting: Dual Engine for Scale-Invariant Self-Supervised Monocular Depth Estimation arxiv
Peizhe Jiang , Wei Yang , Xiaoqing Ye , Xiao Tan , and Meng Wu**
If you find our work useful, please consider citing our paper:
@article{jiang2022detaching,
title={Detaching and Boosting: Dual Engine for Scale-Invariant Self-Supervised Monocular Depth Estimation},
author={Jiang, Peizhe and Yang, Wei and Ye, Xiaoqing and Tan, Xiao and Wu, Meng},
journal={IEEE Robotics and Automation Letters},
volume={7},
number={4},
pages={12094--12101},
year={2022},
publisher={IEEE}
}
Monocular depth estimation (MDE) in the self-supervised scenario has emerged as a promising method as it refrains from the requirement of ground truth depth. Despite continuous efforts, MDE is still sensitive to scale changes especially when all the training samples are from one single camera. Meanwhile, it deteriorates further since camera movement results in heavy coupling between the predicted depth and the scale change. In this paper, we present a scale-invariant approach for self-supervised MDE, in which scale-sensitive features (SSFs) are detached away while scale-invariant features (SIFs) are boosted further. To be specific, a simple but effective data augmentation by imitating camera zooming process is proposed to detach SSFs, making the model robust to scale changes. Besides, a dynamic cross-attention module is designed to boost SIFs by fusing multi-scale cross-attention features adaptively. Extensive experiments on the KITTI dataset demonstrate that the detaching and boosting strategies are mutually complementary in MDE and our approach achieves new State-of-The-Art performance against existing works from 0.097 to 0.090 w.r.t absolute relative error. The code will be made public soon.
We offer two versions of code base: the Paddle version Paddle and the Pytorch version Pytorch , please refer to the specific folders to find the detailed instructments.
