Unsupervised Adaptation For Deep Stereo

Abstract

Recent ground-breaking works have shown that deep neural networks can be trained end-to-end to regress dense disparity maps directly from image pairs. Computer generated imagery is deployed to gather the large data corpus required to train such networks, an additional fine-tuning allowing to adapt the model to work well also on real and possibly diverse environments. Yet, besides a few public datasets such as Kitti, the ground-truth needed to adapt the network to a new scenario is hardly available in practice. In this paper we propose a novel unsupervised adaptation approach that enables to fine-tune a deep learning stereo model without any ground-truth information. We rely on off-the-shelf stereo algorithms together with state-of-the-art confidence measures, the latter able to ascertain upon correctness of the measurements yielded by former. Thus, we train the network based on a novel loss-function that penalizes predictions disagreeing with the highly confident disparities provided by the algorithm and enforces a smoothness constraint. Experiments on popular datasets (KITTI 2012, KITTI 2015 and Middlebury 2014) and other challenging test images demonstrate the effectiveness of our proposal.

BibTex

@inproceedings{tonioni2017unsupervised,
  title={Unsupervised adaptation for deep stereo},
  author={Tonioni, Alessio and Poggi, Matteo and Mattoccia, Stefano and Di Stefano, Luigi},
  booktitle={Proceedings of the IEEE International Conference on Computer Vision},
  pages={1605--1613},
  year={2017}
}