Practical Stereo Matching via Cascaded Recurrent Network with Adaptive Correlation

By Jiankun Li, Peisen Wang, Pengfei Xiong, Tao Cai, Zi-Ping Yan, Lei Yang, Jiangyu Liu, Haoqiang Fan, Shuaicheng Liu. 2022

400 citations recorded by the source corpus.

With the advent of convolutional neural networks, stereo matching algorithms have recently gained tremendous progress. However, it remains a great challenge to accurately extract disparities from real-world image pairs taken by consumer-level devices like smartphones, due to practical complicating factors such as thin structures, non-ideal rectification, camera module inconsistencies and various hard-case scenes. In this paper, we propose a set of innovative designs to tackle the problem of practical stereo matching: 1) to better recover fine depth details, we design a hierarchical network with recurrent refinement to update disparities in a coarse-to-fine manner, as well as a stacked cascaded architecture for inference; 2) we propose an adaptive group correlation layer to mitigate the impact of erroneous rectification; 3) we introduce a new synthetic dataset with special attention to difficult cases for better generalizing to real-world scenes. Our results not only rank 1st on both Middlebury and ETH3D benchmarks, outperforming existing state-of-the-art methods by a notable margin, but also exhibit high-quality details for real-life photos, which clearly demonstrates the efficacy of our contributions.

Subfields: Computer Vision and Pattern Recognition, Model Architectures, 3D Vision & Pose

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Dacheng Tao
2. Jing Zhang
3. Jianfei Cai
4. Luc Van Gool
5. Andreas Geiger
6. Jitendra Malik
7. Ming-Hsuan Yang
8. Xiaogang Wang

In Model Architectures

1. Dacheng Tao
2. Andreas Geiger
3. Jing Zhang
4. Haofei Xu
5. Jianfei Cai
6. Fisher Yu
7. Timm Haucke
8. Volker Steinhage

In 3D Vision & Pose

1. Hjalmar S. Kühl
2. Timm Haucke
3. Volker Steinhage
4. Haofei Xu
5. Dacheng Tao
6. Jianfei Cai
7. Jing Zhang
8. Hamid Rezatofighi

Citations in the corpus

• SOCRATES: Introducing Depth in Visual Wildlife Monitoring Using Stereo Vision (2022, methodology): “Figure 3 shows some exemplary pairs of near infrared images and corresponding depth maps inferred by (Li et al., 2022).”
• Unifying Flow, Stereo and Depth Estimation (2022, methodology): “We note again that the number of refinements required by our final model is much smaller than compared to pure iterative architectures like RAFT-Stereo [26] and CREStereo [62] thanks to our strong discriminative feature representations.”