4 Hits in 1.2 sec

ACTNET: end-to-end learning of feature activations and multi-stream aggregation for effective instance image retrieval [article]

Syed Sameed Husain, Eng-Jon Ong, Miroslaw Bober
2020 arXiv   pre-print
We propose a novel CNN architecture called ACTNET for robust instance image retrieval from large-scale datasets.  ...  This means that our network jointly learns the CNN filters and their optimal activation and aggregation for retrieval tasks.  ...  Visual recognition capabilities are crucial for many applications including mobile visual search, augmented reality, robotic vision, automotive navigation, mobile commerce, surveillance & security, content  ... 
arXiv:1907.05794v3 fatcat:ukmpjk53wvgq3brok354qqwmim

Guest Editorial Special Issue on 6G-Enabled Internet of Things

Qilian Liang, Tariq S. Durrani, Jing Liang, Jinhwan Koh, Xin Wang
2021 IEEE Internet of Things Journal  
Among these applications, intelligent urban surveillance systems for smart cities are becoming more and more important. Therefore, designing a robust visual tracking method has become an urgent task.  ...  In the article, "Multilocation human activity recognition via MIMO-OFDM-based wireless networks: An IoT-inspired device-free sensing approach," an activity decomposition network (ActNet) is presented to  ... 
doi:10.1109/jiot.2021.3111457 fatcat:hs4beids7nh27mbz5ctz7mabvi

Agent-oriented programming

Yoav Shoham
1993 Artificial Intelligence  
The state of an agent consists of components such as beliefs, decisions, capabilities, and obligations; for this reason the state of an agent is called its mental state.  ...  The mental state of agents is described formally in an extension of standard epistemic logics: beside temporalizing the knowledge and belief operators, AOP introduces operators for obligation, decision  ...  ACTNET [58] is a language for computing goalachieving actions that depends dynamically on sensory and stored data. The ACTNET language is based on the concept of action networks [57] .  ... 
doi:10.1016/0004-3702(93)90034-9 fatcat:j7smcjsocfadnbbgsxc2zcc6tm

Online Spatiotemporal Action Detection and Prediction via Causal Representations [article]

Gurkirt Singh
2020 arXiv   pre-print
An action tube is a set of bounding connected over time, which bounds an action instance in space and time.  ...  In this thesis, we focus on video action understanding problems from an online and real-time processing point of view.  ...  Acknowledgements Improving temporal action detection We observed in Chapter 2 in Section 2.2 that proposal based methods for temporal action detection are best at finding temporal boundaries of action  ... 
arXiv:2008.13759v1 fatcat:j6fgbpppkzdjzigtwxjy3zigje