Visible to the public Accurate Deep Net Crowd Counting for Smart IoT Video acquisition devices

TitleAccurate Deep Net Crowd Counting for Smart IoT Video acquisition devices
Publication TypeConference Paper
Year of Publication2020
AuthorsKhadka, A., Argyriou, V., Remagnino, P.
Conference Name2020 16th International Conference on Distributed Computing in Sensor Systems (DCOSS)
Date PublishedMay 2020
ISBN Number978-1-7281-4351-4
Keywordscamera characteristics, Cameras, Computer architecture, consistency, Crowd density estimation, crowd management tasks, crowdsourcing, deep architecture, deep net crowd counting, Deep Neural Network, deep video, Estimation, feature extraction, Internet of Things, Internet of Things cameras, IoT cameras, Kernel, Metrics, neural nets, object detection, pubcrawl, pyramid contextual module, resilience, Resiliency, Robustness, Scalability, security, security system, Self attention network, smart IoT video acquisition devices, video surveillance

A novel deep neural network is proposed, for accurate and robust crowd counting. Crowd counting is a complex task, as it strongly depends on the deployed camera characteristics and, above all, the scene perspective. Crowd counting is essential in security applications where Internet of Things (IoT) cameras are deployed to help with crowd management tasks. The complexity of a scene varies greatly, and a medium to large scale security system based on IoT cameras must cater for changes in perspective and how people appear from different vantage points. To address this, our deep architecture extracts multi-scale features with a pyramid contextual module to provide long-range contextual information and enlarge the receptive field. Experiments were run on three major crowd counting datasets, to test our proposed method. Results demonstrate our method supersedes the performance of state-of-the-art methods.

Citation Keykhadka_accurate_2020