Action Recognition
Action recognition, the task of automatically identifying actions within video data, aims to develop robust and efficient systems for understanding human and animal behavior. Current research focuses on improving accuracy and efficiency across diverse scenarios, employing various model architectures such as transformers, convolutional neural networks, and recurrent neural networks, often incorporating multimodal data (RGB, depth, skeleton, audio) and self-supervised learning techniques. This field is crucial for numerous applications, including autonomous systems, healthcare monitoring, and video surveillance, with ongoing efforts to address challenges like domain generalization, few-shot learning, and adversarial robustness.
Papers
Enhancing Action Recognition by Leveraging the Hierarchical Structure of Actions and Textual Context
Manuel Benavent-Lledo, David Mulero-Pérez, David Ortiz-Perez, Jose Garcia-Rodriguez, Antonis Argyros
Zero-Shot Action Recognition in Surveillance Videos
Joao Pereira, Vasco Lopes, David Semedo, Joao Neves
Human Stone Toolmaking Action Grammar (HSTAG): A Challenging Benchmark for Fine-grained Motor Behavior Recognition
Cheng Liu, Xuyang Yan, Zekun Zhang, Cheng Ding, Tianhao Zhao, Shaya Jannati, Cynthia Martinez, Dietrich Stout
Understanding Spatio-Temporal Relations in Human-Object Interaction using Pyramid Graph Convolutional Network
Hao Xing, Darius Burschka
CHASE: Learning Convex Hull Adaptive Shift for Skeleton-based Multi-Entity Action Recognition
Yuhang Wen, Mengyuan Liu, Songtao Wu, Beichen Ding
Fourier-based Action Recognition for Wildlife Behavior Quantification with Event Cameras
Friedhelm Hamann, Suman Ghosh, Ignacio Juarez Martinez, Tom Hart, Alex Kacelnik, Guillermo Gallego