Robust Stereo Matching Using Discriminative Multilevel Features and Multimodal Bifurcated Cost Volume Network

Okae, James; Li, Bohan; Qin, Huabiao

doi:10.1109/jsen.2023.3246960

Cited by 2 publications

(1 citation statement)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For instance, we can reduce the gap between the target and source domains by matching the distribution of deep features [52,53] or statistics [54,55]. In the area of stereo matching, stereo networks are first trained on synthetic data and then the networks are fine-tuned on real data [56,57]. Although domain adaptation has good results in filling the gap between the synthetic and real domains, rendering synthetic datasets with realistic shapes and various scenes requires significant manual labour and is expensive.…”

Section: Stereo Domain Adaptationmentioning

confidence: 99%

Stereo matching from monocular images using feature consistency

Lu,

Chen,

Gao

et al. 2024

IET Image Processing

View full text Add to dashboard Cite

Synthetic images facilitate stereo matching. However, synthetic images may suffer from image distortion, domain bias, and stereo mismatch, which would significantly restrict the widespread use of stereo matching models in the real world. The first goal in this paper is to synthesize real‐looking images for minimizing the domain bias between the synthesized and real images. For this purpose, sharpened disparity maps are produced from a mono real image. Then, stereo image pairs are synthesized using these imperfect disparity maps and the single real image in the proposed pipeline. Although the synthesized images are as realistic as possible, the domain styles of the synthesized images are always very different from the real images. Thus, the second goal is to enhance the domain generalization ability of the stereo matching network. For that, the feature extraction layer is replaced with a teacher–student model. Then, a constraint of binocular contrast features is imposed on the output of the model. When tested on the KITTI, ETH3D, and Middlebury datasets, the accuracy of the method outperforms traditional methods by at least 30%. Experiments demonstrate that the approaches are general and can be conveniently embedded into existing stereo networks.

show abstract