@inproceedings{930595844efe4d2ca5d113c9fe56fd35,
title = "Human emotion based real-time memory and computation management on resource-limited edge devices",
abstract = "Emotional AI or Affective Computing has been projected to grow rapidly in the upcoming years. Despite many existing developments in the application space, there has been a lack of hardware-level exploitation of the user's emotions. In this paper, we propose a deep collaboration between user's affects and the hardware system management on resource-limited edge devices. Based on classification results from efficient affect classifiers on smartphone devices, novel real-time management schemes for memory, and video processing are proposed to improve the energy efficiency of mobile devices. Case studies on H.264 / AVC video playback and Android smartphone usages are provided showing significant power saving of up to 23% and reduction of memory loading of up to 17% using the proposed affect adaptive architecture and system management schemes.",
keywords = "affective computing, edge devices, LSTM, memory management, system management, wearable devices",
author = "Yijie Wei and Zhiwei Zhong and Jie Gu",
note = "Funding Information: Fig. 10 shows the total amount of loaded memory at App start time and loading time by the selected Apps during the simulated affect-related App usage sequence shown in Fig.9. In the proposed emotion driven management case, the memory loading saving comes from roughly equal saving of file loading from flash drive and app-specific allocated memory space. As shown in the figure, the proposed scheme achieves 17% saving of total memory loaded at App start, and 12% saving of loading time compared to the system default background management scheme. 6 CONCLUSION The real-time emotion based affective computing offers an unprecedented path to tighten the connection between human users and edge devices. In this paper, we explore a deep collaboration between human emotions and embedded hardware management to achieve enhanced computing efficiency on resource-limited edge devices. Different machine learning classifiers were studied to guide the model selection on resource limited edge devices. An affect driven hardware video decoder was proposed achieving up to 23% power saving following the proposed adaptive scheme. An affect driven application and memory management scheme on the Android operating system was also proposed with up to 17% savings on memory loading thanks to the system adaptation to the real-time user{\textquoteright}s affect. ACKNOWLEDGEMENT This work was supported in part by the National Science Foundation under grant number CNS-1816870. Publisher Copyright: {\textcopyright} 2022 ACM.; 59th ACM/IEEE Design Automation Conference, DAC 2022 ; Conference date: 10-07-2022 Through 14-07-2022",
year = "2022",
month = jul,
day = "10",
doi = "10.1145/3489517.3530490",
language = "English (US)",
series = "Proceedings - Design Automation Conference",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "487--492",
booktitle = "Proceedings of the 59th ACM/IEEE Design Automation Conference, DAC 2022",
address = "United States",
}
