提出新型注意力机制，显著提升立体匹配在反光场景下的表现。
Hadamard Attention Recurrent Transformer: A Strong Baseline for Stereo Matching Transformer
- 引入密集注意力核，突破权重上限限制，增强特征交互灵活性。
- 在KITTI 2012反光区域实现第一名，优于提交时所有已发表方法。
- 适合关注立体视觉、注意力机制改进的研究者与工程师。
受注意力机制固有的低秩瓶颈限制，现有立体匹配变换器在非线性表达能力上受限，导致特征表示对反射等挑战性条件敏感。为解决此问题，本文提出哈达玛注意力循环立体变换器（HART）。HART包含一种新注意力机制，包含：1）密集注意力核（DAK），将注意力权重分布映射至(0, +∞)的高维空间，通过移除权重上限约束，实现更灵活的复杂特征交互建模，降低特征共线性；2）多核与阶次交互（MKOI）模块，统一语义与空间知识学习，增强模型对双目图像特征的捕捉能力。实验表明，本方法有效：在提交时所有已发表方法中，于KITTI 2012基准测试的反光区域排名第一。代码已开源。
原文摘要 · Abstract (English)
Constrained by the low-rank bottleneck inherent in attention mechanisms, current stereo matching transformers suffer from limited nonlinear expressivity, which renders their feature representations sensitive to challenging conditions such as reflections. To overcome this difficulty, we present the Hadamard Attention Recurrent Stereo Transformer (HART). HART includes a novel attention mechanism that incorporates the following components: 1) The Dense Attention Kernel (DAK) maps the attention weight distribution into a high-dimensional space over (0, +$\infty$). By removing the upper bound constraint on attention weights, DAK enables more flexible modeling of complex feature interactions. This reduces feature collinearity. 2) The Multi Kernel & Order Interaction (MKOI) module extends the attention mechanism by unifying semantic and spatial knowledge learning. This integration improves the ability of HART to learn features in binocular images. Experimental results demonstrate the effectiveness of our HART. In reflective area, HART ranked 1st on the KITTI 2012 benchmark among all published methods at the time of submission. Code is available at https://github.com/ZYangChen/HART.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。