WINLAB Seminar - Arani Bhattacharya: "Lightweight Sel. of Regions of Interest from Videos for Traf"

Date: July 26, 2024 - 11:00 AM
Title: Lightweight Selection of Regions of Interest from Videos for Traffic Surveillance
Speaker: Prof. Arani Bhattacharya
Abstract: With traffic surveillance increasingly used, thousands of cameras on roads send video feeds to cloud servers to run computer vision algorithms, requiring high bandwidth. State-of-the-art techniques reduce the bandwidth requirement by either sending a limited number of frames/pixels/regions or relying on re-encoding the important parts of the video. This imposes significant overhead on both the camera side and server side compute as re-encoding is expensive.
To mitigate this problem, we propose TILECLIPPER, a system that utilizes tile sampling, where a limited number of rectangular areas within the frames, known as tiles, are sent to the server. TILECLIPPER selects the tiles adaptively by utilizing its correlation with the tile bitrates. We evaluate TILECLIPPER on different datasets having 55 videos in total to show that, on average, our technique reduces ≈ 22% of data sent to the cloud while providing a detection accuracy of 92% with minimal calibration and compute compared to prior works. We show real-time tile filtering of TILECLIPPER even on cheap edge devices like Raspberry Pi 4 and nVidia Jetson Nano. We further show that a live deployment of TILECLIPPER provides over 87% detection accuracy and over 55% bandwidth savings.
Bio: Arani Bhattacharya is an assistant professor at IIIT-Delhi, India. He has completed his PhD from Stony Brook University in 2019 and Masters from Indian Statistical Institute in 2013, all in Computer Science. He works in the field of wireless networks and edge computing. His recent works have focused on emerging applications over wireless networks, such as streaming of 360-degree videos, surveillance of traffic using cameras and running Industry 4.0 applications over wireless networks. In general, his works focus on integrating algorithmic and machine learning techniques into actual systems. His recent works have appeared in venues like IEEE Infocom, USENIX ATC, IEEE/ACM Transactions on Networking and IEEE European Security & Privacy Symposium.