Global Regulation and Excitation via Attention Tuning for Stereo Matching
Stereo matching achieves significant progress with iterative algorithms like RAFT-Stereo and IGEV-Stereo. However, these methods struggle in ill-posed regions with occlusions, textureless, or repetitive patterns, due to a lack of global context and geometric information for effective iterative refinement. To enable the existing iterative approaches to incorporate global context, we propose the Global Regulation and Excitation via Attention Tuning (GREAT) framework which encompasses three attention modules. Specifically, Spatial Attention (SA) captures the global context within the spatial dimension, Matching Attention (MA) extracts global context along epipolar lines, and Volume Attention (VA) works in conjunction with SA and MA to construct a more robust cost-volume excited by global context and geometric details. To verify the universality and effectiveness of this framework, we integrate it into several representative iterative stereo-matching methods and validate it through extensive experiments, collectively denoted as GREAT-Stereo. This framework demonstrates superior performance in challenging ill-posed regions. Applied to IGEV-Stereo, among all published methods, our GREAT-IGEV ranks first on the Scene Flow test set, KITTI 2015, and ETH3D leaderboards, and achieves second on the Middlebury benchmark. Code is available at https://github.com/JarvisLee0423/GREAT-Stereo.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Adaptive Control with Guaranteed Transient Behavior and Zero Steady-State Error for Systems with Time-Varying Parameters
It is nontrivial to achieve global zero-error regulation for uncertain nonlinear systems. The underlying problem becomes even more challenging if mismatched uncertainties and unknown time-varying control gain are involve…
SesameBERT: Attention for Anywhere
Fine-tuning with pre-trained models has achieved exceptional results for many language tasks. In this study, we focused on one such self-attention network model, namely BERT, which has performed well in terms of stacking…
Personalized Semantics Excitation for Federated Image Classification
Federated learning casts a light on the collaboration of distributed local clients with privacy protected to attain a more generic global model. However, significant distribution shift in input/label space across dif…
ClassificationFederated Learningimage-classificationImage Classification+2MR Image Super-Resolution With Squeeze and Excitation Reasoning Attention Network
High-quality high-resolution (HR) magnetic resonance (MR) images afford more detailed information for reliable diagnosis and quantitative image analyses. Deep convolutional neural networks (CNNs) have shown promising…
Image Super-ResolutionSuper-ResolutionA Globally Stable Practically Implementable PI Passivity-based Controller for Switched Power Converters
In this paper we propose a PI passivity-based controller, applicable to a large class of switched power converters, that ensures global state regulation to a desired equilibrium point. A solution to this problem was repo…