通过分层运动对齐提升小物体场景的光流估计精度
HMAFlow: Learning More Accurate Optical Flow via Hierarchical Motion Field Alignment
- 引入分层运动对齐模块与相关性自注意力机制
- 在Sintel和KITTI基准上相对降低14.2%和7.7%误差
- 适合追求高精度光流且关注小目标场景的研究者
光流估计是视觉领域长期存在的基础任务。本文提出HMAFlow方法，针对小物体等复杂场景下的光流估计进行优化。模型核心包含两个组件：分层运动场对齐（HMA）模块和相关性自注意力（CSA）模块。同时，通过多尺度相关性搜索（MCS）层重构4D代价体，并以多范围搜索策略替代传统代价体中的平均池化操作。实验表明，该方法在多个基准上均优于现有最先进方法。相较于RAFT，在Sintel在线基准的clean pass和final pass上分别实现14.2%和3.4%的相对误差降低；在KITTI测试集上，于Fl-all指标上相比RAFT和GMA分别提升6.8%和7.7%。代码将公开于https://github.com/BooTurbo/HMAFlow。
原文摘要 · Abstract (English)
Optical flow estimation is a fundamental and long-standing visual task. In this work, we present a novel method, dubbed HMAFlow, to improve optical flow estimation in challenging scenes, particularly those involving small objects. The proposed model mainly consists of two core components: a Hierarchical Motion Field Alignment (HMA) module and a Correlation Self-Attention (CSA) module. In addition, we rebuild 4D cost volumes by employing a Multi-Scale Correlation Search (MCS) layer and replacing average pooling in common cost volumes with a search strategy utilizing multiple search ranges. Experimental results demonstrate that our model achieves the best generalization performance compared to other state-of-the-art methods. Specifically, compared with RAFT, our method achieves relative error reductions of 14.2% and 3.4% on the clean pass and final pass of the Sintel online benchmark, respectively. On the KITTI test benchmark, HMAFlow surpasses RAFT and GMA in the Fl-all metric by relative margins of 6.8% and 7.7%, respectively. To facilitate future research, our code will be made available at https://github.com/BooTurbo/HMAFlow.
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。