Particle Image Velocimetry (PIV) is a classical flow estimation problem which is widely considered and utilised, especially as a diagnostic tool in experimental fluid dynamics and the remote sensing of environmental flows. Recently, the development of deep learning based methods has inspired new approaches to tackle the PIV problem. These supervised learning based methods are driven by large volumes of data with ground truth training information. However, it is difficult to collect reliable ground truth data in large-scale, real-world scenarios. Although synthetic datasets can be used as alternatives, the gap between the training set-ups and real-world scenarios limits applicability. We present here what we believe to be the first work which takes an unsupervised learning based approach to tackle PIV problems. The proposed approach is inspired by classic optical flow methods. Instead of using ground truth data, we make use of photometric loss between two consecutive image frames, consistency loss in bidirectional flow estimates and spatial smoothness loss to construct the total unsupervised loss function. The approach shows significant potential and advantages for fluid flow estimation. Results presented here demonstrate that our method outputs competitive results compared with classical PIV methods as well as supervised learning based methods for a broad PIV dataset, and even outperforms these existing approaches in some difficult flow cases. Codes and trained models are available at https://github.com/erizmr/UnLiteFlowNet-PIV.
Surrogate-based cross-correlation for particle image velocimetry
School of Mechanical and Electronic Engineering, Wuhan University of Technology（武汉理工大学机械与电子工程学院）；College of Computer Science, Chongqing University（重庆大学计算机学院）；Huazhong University of Science and Technology（华中科技大学）
Pattern-Based Phase-Separation of Tracer and Dispersed Phase Particles in Two-Phase Defocusing Particle Tracking Velocimetry
Institute of Fluids Mechanics (ISTM), Karlsruhe Institute of Technology (KIT), Kaiserstraße 10, 76131 Karlsruhe, Germany（流体力学研究所（ISTM），卡尔斯鲁厄理工学院（KIT））；Karlsruhe Institute of Technology, Institute of Applied and Numerical Mathematics, Karlsruhe, 76133, Germany（卡尔斯鲁厄理工学院，应用与数值数学研究所）
MCFormer: A Multi-Cost-Volume Network and Comprehensive Benchmark for Particle Image Velocimetry
International School Beijing University of Posts and Telecommunications（北京邮电大学国际学院）；College of Engineering Peking University（北京大学工程学院）；College of Physics and Optoelectronic Engineering Harbin Engineering University（哈尔滨工程大学物理与光学电子工程学院）；School of Artificial Intelligence Beijing University of Posts and Telecommunications（北京邮电大学人工智能学院）
Unsupervised single-particle deep clustering via statistical manifold learning
Peking University（北京大学）；Dana-Farber Cancer Institute（丹娜-法伯癌症研究所）；Intel Corporation（英特尔公司）；Peking University Health Science Center（北京大学医学部）；Harvard Medical School（哈佛医学院）
Dual-frame Fluid Motion Estimation with Test-time Optimization and Zero-divergence Loss
University of Chinese Academy of Sciences（中国科学院大学）；AIR, Tsinghua University（清华大学AIR）；Beijing Institute of Technology（北京理工大学）；Beijing Academy of Artificial Intelligence（北京人工智能研究院）
