{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2023,11,24]],"date-time":"2023-11-24T05:38:35Z","timestamp":1700804315381},"reference-count":52,"publisher":"Association for Computing Machinery (ACM)","issue":"6","content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["ACM Trans. Embed. Comput. Syst."],"published-print":{"date-parts":[[2022,11,30]]},"abstract":"\n Recent advances in deep learning have granted unrivaled performance to sensor-based\n human activity recognition (HAR)<\/jats:bold>\n . However, in a real-world scenario, the HAR solution is subject to diverse changes over time such as the need to learn new activity classes or variations in the data distribution of the already-included activities. To solve these issues, previous studies have tried to apply directly the continual learning methods borrowed from the computer vision domain, where it is vastly explored. Unfortunately, these methods either lead to surprisingly poor results or demand copious amounts of computational resources, which is infeasible for the low-cost resource-constrained devices utilized in HAR. In this paper, we provide a resource-efficient and high-performance continual learning solution for HAR. It consists of an expandable neural network trained with a replay-based method that utilizes a highly-compressed replay memory whose samples are selected to maximize data variability. Experiments with four open datasets, which were conducted on two distinct microcontrollers, show that our method is capable of achieving substantial accuracy improvements over baselines in continual learning such as Gradient Episodic Memory, while utilizing only one-third of the memory and being up to 3\u00d7 faster.\n <\/jats:p>","DOI":"10.1145\/3530910","type":"journal-article","created":{"date-parts":[[2022,4,30]],"date-time":"2022-04-30T11:12:57Z","timestamp":1651317177000},"page":"1-25","update-policy":"http:\/\/dx.doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":2,"title":["Resource-Efficient Continual Learning for Sensor-Based Human Activity Recognition"],"prefix":"10.1145","volume":"21","author":[{"ORCID":"http:\/\/orcid.org\/0000-0002-5681-181X","authenticated-orcid":false,"given":"Clayton Frederick Souza","family":"Leite","sequence":"first","affiliation":[{"name":"Aalto University, Espoo, Finland"}]},{"ORCID":"http:\/\/orcid.org\/0000-0002-4517-3779","authenticated-orcid":false,"given":"Yu","family":"Xiao","sequence":"additional","affiliation":[{"name":"Aalto University, Espoo, Finland"}]}],"member":"320","published-online":{"date-parts":[[2022,10,18]]},"reference":[{"key":"e_1_3_1_2_2","doi-asserted-by":"publisher","DOI":"10.1109\/JIOT.2020.3033430"},{"key":"e_1_3_1_3_2","article-title":"Memory aware synapses: Learning what (not) to forget","volume":"1711","author":"Aljundi R.","year":"2017","unstructured":"R. Aljundi, F. Babiloni, M. Elhoseiny, M. Rohrbach, and T. Tuytelaars. 2017. Memory aware synapses: Learning what (not) to forget. CoRR abs\/1711.09601 (2017). arxiv:1711.09601http:\/\/arxiv.org\/abs\/1711.09601.","journal-title":"CoRR"},{"key":"e_1_3_1_4_2","volume-title":"Advances in Neural Information Processing Systems","author":"Aljundi R.","year":"2019","unstructured":"R. Aljundi, M. Lin, B. Goujaud, and Y. Bengio. 2019. Gradient based sample selection for online continual learning. In Advances in Neural Information Processing Systems, H. Wallach, H. Larochelle, A. Beygelzimer, F. d'Alch\u00e9-Buc, E. Fox, and R. Garnett (Eds.), Vol. 32. Curran Associates, Inc.https:\/\/proceedings.neurips.cc\/paper\/2019\/file\/e562cd9c0768d5464b64cf61da7fc6bb-Paper.pdf."},{"key":"e_1_3_1_5_2","doi-asserted-by":"publisher","DOI":"10.3390\/s150613159"},{"key":"e_1_3_1_6_2","volume-title":"International Conference on Learning Representations","author":"Chaudhry A.","year":"2019","unstructured":"A. Chaudhry, M. Ranzato, M. Rohrbach, and M. Elhoseiny. 2019. Efficient lifelong learning with A-GEM. In International Conference on Learning Representations."},{"key":"e_1_3_1_7_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.patrec.2012.12.014"},{"key":"e_1_3_1_8_2","doi-asserted-by":"publisher","DOI":"10.1145\/3447744"},{"key":"e_1_3_1_9_2","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2021.3057446"},{"key":"e_1_3_1_10_2","doi-asserted-by":"publisher","DOI":"10.1145\/3450494"},{"key":"e_1_3_1_11_2","doi-asserted-by":"publisher","DOI":"10.3390\/s19204474"},{"key":"e_1_3_1_12_2","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2018.2846646"},{"key":"e_1_3_1_13_2","unstructured":"I. J. Goodfellow J. Pouget-Abadie M. Mirza B. Xu D. Warde-Farley S. Ozair A. Courville and Y. Bengio. 2014. Generative Adversarial Networks. arxiv:1406.2661 [stat.ML]"},{"key":"e_1_3_1_14_2","doi-asserted-by":"publisher","DOI":"10.1145\/3472290"},{"issue":"2","key":"e_1_3_1_15_2","first-page":"11","article-title":"Ensembles of deep LSTM learners for activity recognition using wearables","volume":"1","author":"Guan Y.","year":"2017","unstructured":"Y. Guan and T. Pl\u00f6tz. 2017. Ensembles of deep LSTM learners for activity recognition using wearables. IMWUT 1, 2, Article 11 (June 2017), 28 pages.","journal-title":"IMWUT"},{"key":"e_1_3_1_16_2","doi-asserted-by":"publisher","DOI":"10.1145\/3372342"},{"key":"e_1_3_1_17_2","first-page":"1533","volume-title":"IJCAI\u201916","author":"Hammerla N. Y.","year":"2016","unstructured":"N. Y. Hammerla, S. Halloran, and T. Pl\u00f6tz. 2016. Deep, convolutional, and recurrent models for human activity recognition using wearables. In IJCAI\u201916. AAAI Press, 1533\u20131540."},{"key":"e_1_3_1_18_2","doi-asserted-by":"publisher","DOI":"10.1109\/MNET.001.1900100"},{"key":"e_1_3_1_19_2","unstructured":"G. Hinton O. Vinyals and J. Dean. 2015. Distilling the Knowledge in a Neural Network. arxiv:1503.02531 [stat.ML]"},{"key":"e_1_3_1_20_2","doi-asserted-by":"crossref","unstructured":"J. Hsiao Y. Deng T. Pao H. Chou and J. Chang. 2017. Design of a wireless 3D hand motion tracking and gesture recognition glove for virtual reality applications Vol. ASME 2017 Conference on Information Storage and Processing Systems. 1\u20133.","DOI":"10.1115\/ISPS2017-5450"},{"key":"e_1_3_1_21_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.ins.2021.04.062"},{"key":"e_1_3_1_22_2","doi-asserted-by":"publisher","DOI":"10.1109\/JSEN.2021.3130761"},{"key":"e_1_3_1_23_2","article-title":"Overcoming catastrophic forgetting in neural networks","volume":"1612","author":"Kirkpatrick J.","year":"2016","unstructured":"J. Kirkpatrick, R. Pascanu, N. C. Rabinowitz, J. Veness, G. Desjardins, A. A. Rusu, K. Milan, J. Quan, T. Ramalho, A. Grabska-Barwinska, D. Hassabis, C. Clopath, D. Kumaran, and R. Hadsell. 2016. Overcoming catastrophic forgetting in neural networks. CoRR abs\/1612.00796 (2016). arxiv:1612.00796http:\/\/arxiv.org\/abs\/1612.00796.","journal-title":"CoRR"},{"key":"e_1_3_1_24_2","doi-asserted-by":"publisher","DOI":"10.1109\/IPSN.2016.7460664"},{"key":"e_1_3_1_25_2","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2020.2993818"},{"key":"e_1_3_1_26_2","doi-asserted-by":"publisher","DOI":"10.1145\/3376897.3377859"},{"key":"e_1_3_1_27_2","doi-asserted-by":"publisher","DOI":"10.1145\/3412382.3458278"},{"key":"e_1_3_1_28_2","article-title":"Learning without forgetting","volume":"1606","author":"Li Z.","year":"2016","unstructured":"Z. Li and D. Hoiem. 2016. Learning without forgetting. CoRR abs\/1606.09282 (2016). arxiv:1606.09282http:\/\/arxiv.org\/abs\/1606.09282.","journal-title":"CoRR"},{"key":"e_1_3_1_29_2","article-title":"Gradient episodic memory for continuum learning","volume":"1706","author":"Lopez-Paz D.","year":"2017","unstructured":"D. Lopez-Paz and M. Ranzato. 2017. Gradient episodic memory for continuum learning. CoRR abs\/1706.08840 (2017). arxiv:1706.08840http:\/\/arxiv.org\/abs\/1706.08840.","journal-title":"CoRR"},{"key":"e_1_3_1_30_2","doi-asserted-by":"publisher","DOI":"10.1109\/TITS.2020.2988504"},{"key":"e_1_3_1_31_2","article-title":"PackNet: Adding multiple tasks to a single network by iterative pruning","volume":"1711","author":"Mallya A.","year":"2017","unstructured":"A. Mallya and S. Lazebnik. 2017. PackNet: Adding multiple tasks to a single network by iterative pruning. CoRR abs\/1711.05769 (2017). arxiv:1711.05769http:\/\/arxiv.org\/abs\/1711.05769.","journal-title":"CoRR"},{"key":"e_1_3_1_32_2","article-title":"Class-incremental learning: Survey and performance evaluation","volume":"2010","author":"Masana M.","year":"2020","unstructured":"M. Masana, X. Liu, B. Twardowski, M. Menta, A. D. Bagdanov, and J. van de Weijer. 2020. Class-incremental learning: Survey and performance evaluation. CoRR abs\/2010.15277 (2020). arxiv:2010.15277https:\/\/arxiv.org\/abs\/2010.15277.","journal-title":"CoRR"},{"key":"e_1_3_1_33_2","article-title":"Spectral normalization for generative adversarial networks","volume":"1802","author":"Miyato T.","year":"2018","unstructured":"T. Miyato, T. Kataoka, M. Koyama, and Y. Yoshida. 2018. Spectral normalization for generative adversarial networks. CoRR abs\/1802.05957 (2018). arxiv:1802.05957http:\/\/arxiv.org\/abs\/1802.05957.","journal-title":"CoRR"},{"key":"e_1_3_1_34_2","doi-asserted-by":"publisher","DOI":"10.3390\/s16010115"},{"key":"e_1_3_1_35_2","article-title":"AHAR: Adaptive CNN for energy-efficient human activity recognition in low-power edge devices","volume":"2102","author":"Rashid N.","year":"2021","unstructured":"N. Rashid, B. U. Demirel, and M. A. Al Faruque. 2021. AHAR: Adaptive CNN for energy-efficient human activity recognition in low-power edge devices. CoRR abs\/2102.01875 (2021). arxiv:2102.01875https:\/\/arxiv.org\/abs\/2102.01875.","journal-title":"CoRR"},{"key":"e_1_3_1_36_2","article-title":"iCaRL: Incremental classifier and representation learning","volume":"1611","author":"Rebuffi S.","year":"2016","unstructured":"S. Rebuffi, A. Kolesnikov, and C. H. Lampert. 2016. iCaRL: Incremental classifier and representation learning. CoRR abs\/1611.07725 (2016). arxiv:1611.07725http:\/\/arxiv.org\/abs\/1611.07725.","journal-title":"CoRR"},{"key":"e_1_3_1_37_2","doi-asserted-by":"publisher","DOI":"10.1109\/ISWC.2012.13"},{"key":"e_1_3_1_38_2","article-title":"Progressive neural networks","volume":"1606","author":"Rusu A. A.","year":"2016","unstructured":"A. A. Rusu, N. C. Rabinowitz, G. Desjardins, H. Soyer, J. Kirkpatrick, K. Kavukcuoglu, R. Pascanu, and R. Hadsell. 2016. Progressive neural networks. CoRR abs\/1606.04671 (2016). arxiv:1606.04671http:\/\/arxiv.org\/abs\/1606.04671.","journal-title":"CoRR"},{"key":"e_1_3_1_39_2","doi-asserted-by":"publisher","DOI":"10.1007\/s12062-020-09260-z"},{"key":"e_1_3_1_40_2","article-title":"Continual learning with deep generative replay","volume":"1705","author":"Shin H.","year":"2017","unstructured":"H. Shin, J. Kwon Lee, J. Kim, and J. Kim. 2017. Continual learning with deep generative replay. CoRR abs\/1705.08690 (2017). arxiv:1705.08690http:\/\/arxiv.org\/abs\/1705.08690.","journal-title":"CoRR"},{"key":"e_1_3_1_41_2","article-title":"Avoiding catastrophic forgetting in mitigating model biases in sentence-pair classification with elastic weight consolidation","volume":"2004","author":"Thorne J.","year":"2020","unstructured":"J. Thorne and A. Vlachos. 2020. Avoiding catastrophic forgetting in mitigating model biases in sentence-pair classification with elastic weight consolidation. CoRR abs\/2004.14366 (2020). arxiv:2004.14366https:\/\/arxiv.org\/abs\/2004.14366.","journal-title":"CoRR"},{"key":"e_1_3_1_42_2","doi-asserted-by":"publisher","DOI":"10.1016\/j.patrec.2018.02.010"},{"key":"e_1_3_1_43_2","doi-asserted-by":"publisher","DOI":"10.1109\/JAS.2021.1004243"},{"key":"e_1_3_1_44_2","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2018.2890675"},{"key":"e_1_3_1_45_2","article-title":"DFTerNet: Towards 2-bit dynamic fusion networks for accurate human activity recognition","volume":"1808","author":"Yang Z.","year":"2018","unstructured":"Z. Yang, O. Ian Raymond, C. Zhang, Y. Wan, and J. Long. 2018. DFTerNet: Towards 2-bit dynamic fusion networks for accurate human activity recognition. CoRR abs\/1808.04228 (2018). arxiv:1808.04228http:\/\/arxiv.org\/abs\/1808.04228.","journal-title":"CoRR"},{"key":"e_1_3_1_46_2","doi-asserted-by":"publisher","DOI":"10.3233\/AIS-200566"},{"key":"e_1_3_1_47_2","doi-asserted-by":"publisher","DOI":"10.1145\/3440036"},{"key":"e_1_3_1_48_2","doi-asserted-by":"publisher","DOI":"10.5555\/1786014.1786017"},{"key":"e_1_3_1_49_2","doi-asserted-by":"publisher","DOI":"10.1145\/3267242.3267286"},{"key":"e_1_3_1_50_2","article-title":"Improved multitask learning through synaptic intelligence","volume":"1703","author":"Zenke F.","year":"2017","unstructured":"F. Zenke, B. Poole, and S. Ganguli. 2017. Improved multitask learning through synaptic intelligence. CoRR abs\/1703.04200 (2017). arxiv:1703.04200http:\/\/arxiv.org\/abs\/1703.04200.","journal-title":"CoRR"},{"key":"e_1_3_1_51_2","doi-asserted-by":"publisher","DOI":"10.1109\/TMC.2021.3062575"},{"key":"e_1_3_1_52_2","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2019.2920969"},{"key":"e_1_3_1_53_2","doi-asserted-by":"publisher","DOI":"10.1109\/JSEN.2021.3122258"}],"container-title":["ACM Transactions on Embedded Computing Systems"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3530910","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,1,1]],"date-time":"2023-01-01T19:19:33Z","timestamp":1672600773000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3530910"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022,10,18]]},"references-count":52,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2022,11,30]]}},"alternative-id":["10.1145\/3530910"],"URL":"https:\/\/doi.org\/10.1145\/3530910","relation":{},"ISSN":["1539-9087","1558-3465"],"issn-type":[{"value":"1539-9087","type":"print"},{"value":"1558-3465","type":"electronic"}],"subject":[],"published":{"date-parts":[[2022,10,18]]},"assertion":[{"value":"2021-07-15","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-04-08","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2022-10-18","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}
