Overview
Understanding human activity from images and videos is one of the main goals of computer vision-based systems. Over the years, VRL researchers have developed algorithms to determine complex human activity in different modalities. In the past, the researchers from the lab have collaborated with various organizations in creating different systems and algorithms including an online system to detect atomic human activities from continuous video streams, person reidentification and activity detection algorithms etc. In recent years, researchers in the VRL lab have focussed to develop attention-based deep neural networks to understand human-object interactions. Moreover, researchers are working to evolve already developed activity detection algorithms in understanding complex human activities that span over multiple cameras or a combination of multiple atomic actions.
Related Publications
“Actor Conditioned Attention Maps for Video Action Detection”, Oytun Ulutan, Swati Rallapalli, Mudhakar Srivatsa, B.S. Manjunath, The IEEE Winter Conference on Applications of Computer Vision,2020,527--536, Colorado, USA, Jan. 2020.
“A Multiview Multimodal System for Monitoring Patient Sleep”, C. Torres, J. C. Fried, K. Rose and B. S. Manjunath,
IEEE Tran. Multimedia, May. 2018.
“A Bilinear Model for Person Detection in Multi-Modal Data”, Oytun Ulutan, Benjamin Riggan, Nasser Nasrabadi, B.S. Manjunath, IEEE Winter Conf. on Applications of Computer Vision (WACV 2018), March 12-14, 2018, Lake Tahoe, NV/CA, Lake Tahoe, CA, Mar. 2018.
“KESTREL: Video Analytics for Augmented Multi-Camera Vehicle Tracking”, Hang Qiu, Xiaochen Liu, Swati Rallapalli, Archith J. Bency, Rahul Urgaonkar, B.S. Manjunath, Ramesh Govindan, Kevin Chan, The Proceedings of ACM/IEEE International Conference on Internet-of-Things Design and Implementation (IoTDI), Orlando, Florida, Apr. 2018.
