Flow Mood
"Flow," in various scientific contexts, refers to the dynamic movement or transformation of data or processes, often aiming to improve efficiency, accuracy, or personalization. Current research focuses on leveraging flow-based models, including normalizing flows, rectified flows, and neural ordinary differential equations, to address challenges in diverse fields such as image generation, video analysis, and robotics. These advancements are significantly impacting areas like automated vehicle validation, human activity recognition, and high-energy physics simulations by enabling more efficient and accurate modeling of complex systems and data.
Papers
Stable Flow: Vital Layers for Training-Free Image Editing
Omri Avrahami, Or Patashnik, Ohad Fried, Egor Nemchinov, Kfir Aberman, Dani Lischinski, Daniel Cohen-Or
FLRNet: A Deep Learning Method for Regressive Reconstruction of Flow Field From Limited Sensor Measurements
Phong C. H. Nguyen, Joseph B. Choi, Quang-Trung Luu