LIMU-BERT
The DOI matches its Crossref record; 54 of the 56 references it cites were found in a registry or an index.
Abstract
Deep learning greatly empowers Inertial Measurement Unit (IMU) sensors for various mobile sensing applications, including human activity recognition, human-computer interaction, localization and tracking, and many more. Most existing works require substantial amounts of well-curated labeled data to train IMU-based sensing models, which incurs high annotation and training costs. Compared with labeled data, unlabeled IMU data are abundant and easily accessible. In this work, we present LIMU-BERT, a novel representation learning model that can make use of unlabeled IMU data and extract generalized rather than task-specific features. LIMU-BERT adopts the principle of self-supervised training of the natural language model BERT to effectively capture temporal relations and feature distributions in IMU sensor measurements. However, the original BERT is not adaptive to mobile IMU data. By meticulously observing the characteristics of IMU sensors, we propose a series of techniques and accordingly adapt LIMU-BERT to IMU sensing tasks. The designed models are lightweight and easily deployable on mobile devices. With the representations learned via LIMU-BERT, task-specific models trained with limited labeled samples can achieve superior performances. We extensively evaluate LIMU-BERT with four open datasets. The results show that the LIMU-BERT enhanced models significantly outperform existing approaches in two typical IMU sensing applications.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“On Opportunity and Realdisp datasets, LIMU-BERT [71] achieves the best performance.”
From Spatial-Temporal Masked Autoencoder for Multi-Device Wearable Human Activity Recognition (2023) · cites this paper as [71]
Citations
Cites
Hochreiter, S., & Schmidhuber, J. (1997) · Neural Computation · doi:10.1162/neco.1997.9.8.1735 · Cited by 102,318
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805 · Cited by 73,173
(2008) · Journal of Machine Learning Research · Cited by 33,051
A survey on Image Data Augmentation for Deep Learning
Shorten, C., & Khoshgoftaar, T. M. (2019) · Journal of Big Data · doi:10.1186/s40537-019-0197-0 · Cited by 13,203
(2017) · Cited by 11,625
Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
Chung, J., Gulcehre, C., Cho, K., & Bengio, Y. (2014) · arXiv · doi:10.48550/arxiv.1412.3555 · Cited by 11,008
Extracting and composing robust features with denoising autoencoders
Vincent, P., Larochelle, H., Bengio, Y., & Manzagol, P.-A. (2008) · Proceedings of the 25th international conference on Machine learning - ICML '08 · doi:10.1145/1390156.1390294 · Cited by 7,436
ALBERT: A Lite BERT for Self-supervised Learning of Language Representations
Lan, Z., Chen, M., Goodman, S., Gimpel, K., Sharma, P., & Soricut, R. (2019) · arXiv · doi:10.48550/arxiv.1909.11942 · Cited by 4,308 · cited 3×
“Second, LIMU-BERT adopts a cross-layer parameter sharing mechanism [19] to improve parameter efficiency.”
BackgroundMethodThe class imbalance problem: A systematic study1
Japkowicz, N., & Stephen, S. (2002) · Intelligent Data Analysis · doi:10.3233/ida-2002-6504 · Cited by 3,295
A survey on vision-based human action recognition
Poppe, R. (2010) · Image and Vision Computing · doi:10.1016/j.imavis.2009.11.014 · Cited by 2,242
Research works printed but not linked: 8
- Jimmy Lei Ba , Jamie Ryan Kiros, and Geoffrey E. Hinton . 2016 . Layer Normalization . arXiv:1607.06450 [stat.ML] Jimmy Lei Ba, Jamie Ryan Kiros, and Geoffrey E. Hinton. 2016. Layer Normalization. arXiv:1607.06450 [stat.ML] · arXiv
- Dan Hendrycks and Kevin Gimpel . 2016. Gaussian error linear units (gelus). arXiv preprint arXiv:1606.08415 ( 2016 ). Dan Hendrycks and Kevin Gimpel. 2016. Gaussian error linear units (gelus). arXiv preprint arXiv:1606.08415 (2016). · arXiv
- Diederik P Kingma and Jimmy Ba . 2014 . Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014). Diederik P Kingma and Jimmy Ba. 2014. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014). · arXiv
- Jakub Konečny , H Brendan McMahan , Felix X Yu , Peter Richtárik , Ananda Theertha Suresh, and Dave Bacon . 2016 . Federated learning: Strategies for improving communication efficiency. arXiv preprint arXiv:1610.05492 (2016). Jakub Konečny, H Brendan McMahan, Felix X Yu, Peter Richtárik, Ananda Theertha Suresh, and Dave Bacon. 2016. Federated learning: Strategies for improving communication efficiency. arXiv preprint arXiv:1610.05492 (2016). · arXiv
- Alex Krizhevsky , Ilya Sutskever , and Geoffrey E Hinton . 2012. Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems 25 ( 2012 ), 1097--1105. Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. 2012. Imagenet classification with deep convolutional neural networks. Advances in neural information processing systems 25 (2012), 1097--1105.
- Brendan McMahan Eider Moore Daniel Ramage Seth Hampson and Blaise Aguera y Arcas. 2017. Communication-efficient learning of deep networks from decentralized data. In Artificial intelligence and statistics. PMLR 1273--1282. Brendan McMahan Eider Moore Daniel Ramage Seth Hampson and Blaise Aguera y Arcas. 2017. Communication-efficient learning of deep networks from decentralized data. In Artificial intelligence and statistics. PMLR 1273--1282.
- Ashish Vaswani , Noam Shazeer , Niki Parmar , Jakob Uszkoreit , Llion Jones , Aidan N Gomez , Lukasz Kaiser , and Illia Polosukhin . 2017. Attention is all you need. arXiv preprint arXiv:1706.03762 ( 2017 ). Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. arXiv preprint arXiv:1706.03762 (2017). · arXiv
- Jianbo Yang , Minh Nhut Nguyen , Phyo Phyo San , Xiaoli Li , and Shonali Krishnaswamy . 2015 . Deep convolutional neural networks on multichannel time series for human activity recognition .. In Ijcai , Vol. 15 . Buenos Aires, Argentina, 3995--4001. Jianbo Yang, Minh Nhut Nguyen, Phyo Phyo San, Xiaoli Li, and Shonali Krishnaswamy. 2015. Deep convolutional neural networks on multichannel time series for human activity recognition.. In Ijcai, Vol. 15. Buenos Aires, Argentina, 3995--4001.
Cited by
Cosmo: contrastive fusion learning with small data for multimodal human activity recognition
Ouyang, X., Shuai, X., Zhou, J., Shi, I. W., Xie, Z., Xing, G., & Huang, J. (2022) · Proceedings of the 28th Annual International Conference on Mobile Computing And Networking · doi:10.1145/3495243.3560519 · Cited by 94
“Machine learning algorithms based on handcrafted features [24, 47] and deep neural networks [25, 54] have been applied to HAR applications.”
BackgroundMethodInfluentialYin, Y., Xie, L., Jiang, Z., Xiao, F., Cao, J., & Lu, S. (2024) · IEEE Communications Surveys & Tutorials · doi:10.1109/comst.2024.3357591 · Cited by 108
“When combining multi-modal sensor data, different recognition approaches were also used [135], e.g., using image processing to track fingers in 2D space, while using phase changes of ultrasound to track the depth of fingers in 3D space, to provide a depth-aware tapping scheme.”
Artificial Intelligence of Things: A Survey
Siam, M. S. I., Ahn, H., Liu, L., Alam, S., Shen, H., Cao, Z., Shroff, N. B., Krishnamachari, B., Srivastava, M. B., & Zhang, M. (2024) · ACM Transactions on Sensor Networks · doi:10.1145/3690639 · Cited by 123
Harmony: Heterogeneous Multi-Modal Federated Learning through Disentangled Model Training
Ouyang, X., Xie, Z., Fu, H., Cheng, S., Pan, L., Ling, N., Xing, G., Zhou, J., & Huang, J. (2023) · International Conference on Mobile Systems, Applications, and Services · doi:10.1145/3581791.3596844 · Cited by 109
“Previous studies on semi-supervised learning have shown that the model performance can be augmented by efficiently learning from the unlabeled multi-modal [42] or unimodal [23, 61] data.”
BackgroundHong, Z., Li, Z., Zhong, S., Lyu, W., Wang, H., Ding, Y., He, T., & Zhang, D. (2024) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3659597 · Cited by 67
“Even with good performance, most pretraining methods are not designed for IMU sensor data or not designed for the generalization setting [83].”
BackgroundMethodInfluentialPenetrative AI: Making LLMs Comprehend the Physical World
Xu, H., Han, L., Yang, Q., Li, M., & Srivastava, M. B. (2024) · IEEE Workshop on Mobile Computing Systems and Applications (WMCSA) · doi:10.1145/3638550.3641130 · Cited by 147
“Developing customized and highly effective natural language algorithms for LLMs represents another promising future avenue.”
Self-citationPractically Adopting Human Activity Recognition
Xu, H., Zhou, P., Tan, R., & Li, M. (2023) · Proceedings of the 29th Annual International Conference on Mobile Computing and Networking · doi:10.1145/3570361.3613299 · Cited by 55
Self-citationOuyang, X., Shuai, X., Li, Y., Pan, L., Zhang, X., Fu, H., Cheng, S., Wang, X., Cao, S., Xin, J., et al. (2024) · ACM Annual International Conference on Mobile Computing and Networking (MobiCom) · doi:10.1145/3636534.3649370 · Cited by 86
A Simple Self-Supervised IMU Denoising Method for Inertial Aided Navigation
Yuan, K., & Wang, Z. J. (2023) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2023.3234778 · Cited by 38
“A naive implementation of LIMU-BERT on IMU denosing would be as follows: However, it should be noted that LIMU-BERT [18] is only used to learn a good IMU presentation instead of explicitly doing IMU denosing.”
BackgroundMethodInfluentialEdgeFM: Leveraging Foundation Model for Open-set Learning on the Edge
Yang, B., He, L., Ling, N., Yan, Z., Xing, G., Shuai, X., Ren, X., & Jiang, X. (2023) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3625687.3625793 · Cited by 44
“Most current on-device AI models are task-specific and can only predict a closed-set of classes pre-defined at the training stage [42, 62].”
BackgroundHaresamudram, H., Tang, C. I., Suh, S., Lukowicz, P., & Plötz, T. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3729467 · Cited by 33
“…recognition is yet to be verified, these works provide a good vision of how foundational models for human activity recognition can be: that training with a well-selected set of datasets and modest model size could offer much more practical and effective solutions for this mobile sensing task.”
Human Activity Recognition Using Wearable Sensors Based on Image Classification
Zebhi, S. (2022) · IEEE Sensors Journal · doi:10.1109/jsen.2022.3174280 · Cited by 36
FedConv: A Learning-on-Model Paradigm for Heterogeneous Federated Clients
Shen, L., Yang, Q., Cui, K., Zheng, Y., Wei, X.-Y., Liu, J., & Han, J. (2024) · International Conference on Mobile Systems, Applications, and Services · doi:10.1145/3643832.3661880 · Cited by 32
Federated Few-Shot Learning for Mobile NLP
Cai, D., Wang, S., Wu, Y., Lin, F. X., & Xu, M. (2023) · ACM Annual International Conference on Mobile Computing and Networking (MobiCom) · doi:10.1145/3570361.3613277 · Cited by 44
“, to tag which words of a text message belong to named identities) [93] or the knowledge to do so (e.”
BackgroundOnline Semi-Supervised Transformer for Resilient Vehicle GNSS/INS Navigation
Wang, H., Tang, F., Wei, J., Zhu, B., Wang, Y., & Zhang, K. (2024) · IEEE Transactions on Vehicular Technology · doi:10.1109/tvt.2024.3422680 · Cited by 33
Oleh, U., Obermaisser, R., & Ahammed, A. S. (2024) · Algorithms · doi:10.3390/a17100434 · Cited by 38
Liu, Y., Wang, H., Man, F., Xu, J., Dang, F., Liu, Y., Zhang, X.-P. S., & Chen, X. (2024) · International Conference on Mobile Systems, Applications, and Services · doi:10.1145/3643832.3661872 · Cited by 33
“The original MLM task in BERT masks only one token in a text sequence since many words have independent meanings [38].”
MethodConstruct 3D Hand Skeleton with Commercial WiFi
Ji, S., Zhang, X., Zheng, Y., & Li, M. (2023) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3625687.3625812 · Cited by 40
Self-citationSelf-supervised Learning for Accelerometer-based Human Activity Recognition: A Survey
Logacjov, A. (2024) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3699767 · Cited by 40
Spatial-Temporal Masked Autoencoder for Multi-Device Wearable Human Activity Recognition
Miao, S., Chen, L., & Hu, R. (2023) · Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies · doi:10.1145/3631415 · Cited by 23
“Masked reconstruction tasks leverage encoders to extract valid features from partially masked input and decoders to recover the complete input by learning the correlation patterns within the data [25, 56, 71].”
BackgroundMethodResultInfluentialJawthenticate: Microphone-free Speech-based Authentication using Jaw Motion and Facial Vibrations
Srivastava, T., Pan, S., Nguyen, P., & Jain, S. (2023) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3625687.3625813 · Cited by 30
AlMuhaideb, S., AlAbdulkarim, L., AlShahrani, D. M., AlDhubaib, H., & AlSadoun, D. E. (2024) · Sensors · doi:10.3390/s24165436 · Cited by 32
MMTSA: Multi-Modal Temporal Segment Attention Network for Efficient Human Activity Recognition
Gao, Z., Wang, Y., Chen, J., Xing, J., Patel, S., Liu, X., & Shi, Y. (2023) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3610872 · Cited by 44
“In previous work, a standard input method is to feed the collected data series into deep models [50, 58], such as CNN, RNN, LSTM, or Bert.”
MethodDTMM: Deploying TinyML Models on Extremely Weak IoT Devices with Pruning
Han, L., Xiao, Z., & Li, Z. (2024) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom52122.2024.10621387 · Cited by 34
SwapNet: Efficient Swapping for DNN Inference on Edge AI Devices Beyond the Memory Budget
Wang, K., Cao, J., Zhou, Z., & Li, Z. (2024) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2024.3355764 · Cited by 27
“We first demonstrate the limited memory budget to run typical DNN inference tasks on standard edge AI devices via a measurement study (Section II-A) and then highlight the uniqueness when designing memory-friendly DNN inference strategies on edge AI devices compared with mobile devices (Section…”
MethodFreqMAE: Frequency-Aware Masked Autoencoder for Multi-Modal IoT Sensing
Kara, D., Kimura, T., Liu, S., Li, J., Liu, D., Wang, T., Wang, R., Chen, Y., Hu, Y., & Abdelzaher, T. (2024) · International World Wide Web Conference (WWW) · doi:10.1145/3589334.3645346 · Cited by 38
The role of Large Language Models in addressing IoT challenges: A systematic literature review
Vito, G. D., Palomba, F., & Ferrucci, F. (2025) · Future Generation Computer Systems · doi:10.1016/j.future.2025.107829 · Cited by 30
Robust Route Planning under Uncertain Pickup Requests for Last-mile Delivery
Yan, H., Tan, H., Wang, H., Zhang, D., & Yang, Y. (2024) · International World Wide Web Conference (WWW) · doi:10.1145/3589334.3645595 · Cited by 30
Effective Heterogeneous Federated Learning via Efficient Hypernetwork-based Weight Generation
Shin, Y., Lee, K., Lee, S., Choi, Y. R., Kim, H., & Ko, J. (2024) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3666025.3699326 · Cited by 33
“In contrast, the feature extractor (i.e., convolutional layers) focuses on extracting universally applicable features across client models [18, 45, 69], maximizing the benefits of weight generation.”
LLMs and IoT: A Comprehensive Survey on Large Language Models and the Internet of Things
Sarhaddi, F., Nguyen, N. T., Zuniga, A., Khan, M., Hui, P., Tarkoma, S., Flores, H., & Nurmi, P. (2026) · IEEE Open Journal of the Communications Society · doi:10.1109/ojcoms.2026.3680064 · Cited by 33
“Multimodal LLMs like LIMU-BERT [243] and Llama [26] can interpret sensor data.”
Automatic Update for Wi-Fi Fingerprinting Indoor Localization via Multi-Target Domain Adaptation
Wang, J., Zhao, Z., Ou, M., et al. (2023) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3596239 · Cited by 15
“In order to associate crowdsourced fingerprints with locations, many researches leverage inertial sensors or floorplans to assist localizing, yet introducing extra costs [29, 41, 47] .”
BackgroundNeural-Kalman GNSS/INS Navigation for Precision Agriculture
Du, Y., Saha, S. S., Sandha, S. S., Lovekin, A., Wu, J., Siddharth, S., Chowdhary, M., Jawed, M. K., & Srivastava, M. B. (2023) · Proceedings - IEEE International Conference on Robotics and Automation/Proceedings · doi:10.1109/icra48891.2023.10161351 · Cited by 32
GOAT: A Generalized Cross-Dataset Activity Recognition Framework with Natural Language Supervision
Miao, S., & Chen, L. (2024) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3699736 · Cited by 27
Xu, J., Li, D., Yang, Z., Zhao, Y., Cao, H., Liu, Y., & Shangguan, L. (2023) · ACM Annual International Conference on Mobile Computing and Networking (MobiCom) · doi:10.1145/3570361.3613269 · Cited by 23
Sun, K., Xia, C., Zhang, X., Chen, H., & Zhang, J. (2024) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3643553 · Cited by 21
“SSL has been introduced recently to train feature representation models based on motion sensor data [24, 36–38].”
BackgroundMethodInfluentialMESEN: Exploit Multimodal Data to Design Unimodal Human Activity Recognition with Few Labels
Xu, L., Gu, C., Tan, R., He, S., & Chen, J. (2023) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3625687.3625782 · Cited by 26
“Depending on the modalities involved in the model design and deployment phases, existing studies can be divided into the categories of single-modality [31, 33] and multi-modality [10, 14, 23, 26, 42].”
BackgroundMethodInfluentialSelf-citationR-AFNIO: Redundant IMU fusion with attention mechanism for neural inertial odometry
Yang, B. S., Wang, X., Huang, F., Cao, X., & Zhang, Z. (2024) · Expert Systems with Applications · doi:10.1016/j.eswa.2024.125894 · Cited by 13
LLM4HAR: Generalizable On-device Human Activity Recognition with Pretrained LLMs
Hong, Z., Song, Y., Li, Z., Yu, A., Zhong, S., Ding, Y. C., He, T., & Zhang, D. (2025) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3711896.3737226 · Cited by 34
Zhou, P., Ding, Y. C., Li, Y., Li, M., Shen, G., & He, T. (2022) · ACM Annual International Conference on Mobile Computing and Networking (MobiCom) · doi:10.1145/3495243.3517023 · Cited by 12
“We expect future studies on how traditional crowdsourcing solutions may be complemented with unsupervised or self-supervised learning technology [35] to deal with the scale of unlabeled data that we face.”
BackgroundSelf-citationZhang, J., Feng, L., Liu, Z., Wu, Y., He, Y., Dong, Y., & Xu, D. (2024) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3637528.3671828 · Cited by 18
CatUA: Catalyzing Urban Air Quality Intelligence Through Mobile Crowd-Sensing
Zhou, N., Liu, Y., Wang, H., Man, F., Xu, J., Dang, F., Hong, C., Liu, Y., Zhang, X.-P. S., Song, Y., et al. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3560120 · Cited by 18
“The original MLM task in BERT masks only one token in a text sequence since many words have independent meanings [33].”
PhyMask: An Adaptive Masking Paradigm for Efficient Self-Supervised Learning in IoT
Kara, D., Kimura, T., Chen, Y., Li, J., Wang, R., Chen, Y., Wang, T., Liu, S., & Abdelzaher, T. (2024) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3666025.3699325 · Cited by 20
Advancing PPG-Based Continuous Blood Pressure Monitoring from a Generative Perspective
Ji, H., & Zhou, P. (2024) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3666025.3699365 · Cited by 14
“In addition to the PPG and ECG, several other features have been utilized for estimating BP, such as Pulse Transit Time (PTT) [20, 23, 55, 70], Pulse Wave Velocity (PWV) [26, 53, 74], and Ballisto-cardiography (BCG) [12, 47, 68].”
Self-citationContrastSense: Domain-invariant Contrastive Learning for In-the-Wild Wearable Sensing
Dai, G., Xu, H., Yoon, H., Li, M., Tan, R., & Lee, S.-J. (2024) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3699744 · Cited by 27
Self-citationComparing self-supervised learning techniques for wearable human activity recognition
Ek, S., Presotto, R., Civitarese, G., Portet, F., Lalanda, P., & Bettini, C. (2025) · CCF Transactions on Pervasive Computing and Interaction · doi:10.1007/s42486-024-00182-9 · Cited by 15
“1, most of these works have focused on contrastive approaches (Saeed et al. 2020; Khaertdinov et al. 2021; Qian et al. 2022; Wang et al. 2022; Jain et al. 2022; Zhang et al. 2022; Arrotta et al. 2023; Taghanaki et al. 2023; Meng et al. 2023).”
mm-Fall: Practical and Robust Fall Detection via mmWave Signals
Zhao, C., Luo, Q., Ding, H., Wang, G., Zhao, K., Wang, Z., Xi, W., & Zhao, J. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3557504 · Cited by 15
Masked Video and Body-Worn IMU Autoencoder for Egocentric Action Recognition
Zhang, M., Huang, Y., Liu, R., & Sato, Y. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-72649-1_18 · Cited by 31
Lightweight human activity recognition method based on the MobileHARC model
Gong, X., Zhang, X., & Li, N. (2024) · Systems Science & Control Engineering · doi:10.1080/21642583.2024.2328549 · Cited by 13
Gorges, T., Davidson, P., Boeschen, M., Hotho, A., & Merz, C. (2024) · Sensors · doi:10.3390/s24216773 · Cited by 12
“A promising approach to improve airtime detection is the use of inertial measurement units (IMUs), which have already been widely used for human activity recognition (HAR) tasks [8–10].”
BackgroundFacilitating Radar-Based Gesture Recognition With Self-Supervised Learning
Sheng, Z., Xu, H., Zhang, Q., & Wang, D. (2022) · IEEE International Conference on Sensing, Communication, and Networking (SECON) · doi:10.1109/secon55815.2022.9918549 · Cited by 16
“To this end, we adopt the key principle of self-supervised learning [13]–[16] and propose a novel representations learning framework, RadarAE.”
BackgroundMethodSelf-citationLightHART: Lightweight Human Activity Recognition Transformer
Haque, S. T., Ni, J., Li, J., Yan, Y., & Ngu, A. H. H. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-78354-8_27 · Cited by 10
Deep Networks and Sensor Fusion for Personal Care Robot Tasks—A Review
Ramsai, N., & Sridharan, K. (2025) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3530007 · Cited by 8
“Similarly, another variant of BERT, named lightweight IMU-BERT [71], is used for HAR and achieves an accuracy of 92.7%.”
InfluentialMobHAR: Source-free Knowledge Transfer for Human Activity Recognition on Mobile Devices
Xue, M., Zhu, Y., Xie, W., Wang, Z., Chen, Y., Jiang, K., & Zhang, Q. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3712620 · Cited by 14
“Existing HAR models such as [17, 19, 21, 27, 40, 45, 46, 56, 57] commonly assume that data adheres to the concept of independent and identically distributed (IID) [50].”
MASTER: A Multi-modal Foundation Model for Human Activity Recognition
Zhu, G., Zhao, D. Y., Li, C., Zhao, M., Zhang, Z., Quan, H., & Ma, H. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3749511 · Cited by 16
Disentangling rodent behaviors to improve automated behavior recognition
van Dam, E. A., Noldus, L. P. J. J., & van Gerven, M. A. J. (2023) · Frontiers in Neuroscience · doi:10.3389/fnins.2023.1198209 · Cited by 13
Unvoiced: Designing an LLM-assisted Unvoiced User Interface using Earables
Srivastava, T., Khanna, P., Pan, S., Nguyen, P., & Jain, S. (2024) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3666025.3699374 · Cited by 19
“• LIMU-BERT (IMU Modelling Based): We use their pre-trained model [99] to extract representations, followed by a bidirectional LSTM head serving as a classifier.”
Exploring the Capabilities of LLMs for IMU-based Fine-grained Human Activity Understanding
Xu, L., Hou, K., & Jiang, X. (2025) · doi:10.1145/3722565.3727195 · Cited by 14
Babel: A Scalable Pre-trained Model for Multi-Modal Sensing via Expandable Modality Alignment
Dai, S., Jiang, S., Yang, Y., Cao, T., Li, M., Banerjee, S., & Qiu, L. (2025) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3715014.3722068 · Cited by 26
Self-citationEnhancing Sensor-Based Human Activity Recognition With Multiresolution Wavelet-Attention
Wieland, C., & Pankratius, V. (2025) · IEEE Sensors Journal · doi:10.1109/jsen.2025.3571572 · Cited by 14
“Common pretraining objectives include signal reconstruction [36], masked signal modeling [37], signal forecasting [38], transformation detection [39], and contrastive learning [40].”
UltraCLR: Contrastive Representation Learning Framework for Ultrasound-based Sensing
Wang, X., Yang, Z., Wang, W., Dai, H., Shi, S., & Gu, Q. (2023) · ACM Transactions on Sensor Networks · doi:10.1145/3597498 · Cited by 8
“Recent works have recognized the value of using massive amounts of unlabeled data and proposed self-supervised learning frameworks for Inertial Measurement Units (IMU) based HAR [4, 24, 31, 40].”
Predicting Adverse Events for Patients with Type-1 Diabetes Via Self-Supervised Learning
Zheng, X., Ji, S., & Wu, C. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10446832 · Cited by 15
“After that, span-mask [14] is implemented to mask the monitoring data x at selected time stamps.”
MethodBeyond "Taming Electric Scooters": Disentangling Understandings of Micromobility Naturalistic Riding
Tabatabaie, M., He, S., Wang, H., & Shin, K. G. (2024) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3678513 · Cited by 10
“…[9]: adapted from CLIP and augmented with bidirectional attention mechanism; • FLAVA [58]: with the visual and textual fusion architecture; • LIMU [83]: which applies the BERT [13] mechanism on behavioral data; • TS2Vec [85]: with contrastive learning on behaviors; and • TFC [87]: focusing on…”
MethodAutoLife: Automatic Life Journaling with Smartphones and LLMs
Xu, H., Zeng, Z., Tong, P., Li, M., & Srivastava, M. B. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3770683 · Cited by 26
Self-citationNarayana, M. V., Rachavarapu, K. K., Jalihal, D., & Nagendra, S. M. S. (2024) · IEEE Sensors Journal · doi:10.1109/jsen.2024.3362962 · Cited by 7
“Restrictions apply. of samples in X i using span masking technique [47] with probability ( P ) of 0.2.”
MethodInfluentialA Spatio-temporal Graph Transformer driven model for recognizing fine-grained data human activity
Yan, M., Zhang, G., & Ye, C. (2024) · Alexandria Engineering Journal · doi:10.1016/j.aej.2024.06.042 · Cited by 7
Self-Supervised Learning for Activity Recognition Based on Datasets With Imbalanced Classes
Hamad, R. A., Woo, W. L., Wei, B. W., & Yang, L. (2025) · IEEE Transactions on Emerging Topics in Computational Intelligence · doi:10.1109/tetci.2025.3526504 · Cited by 11
Detecting Cyber Attacks in Smart Grids with Massive Unlabeled Sensing Data
Zeng, H., Ng, Z. W., Zhou, P., Lou, X., Yau, D. K. Y., & Winslett, M. (2022) · IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm) · doi:10.1109/smartgridcomm52983.2022.9961017 · Cited by 7
Self-citationChang, K. C., Hasegawa‐Johnson, M., McElwain, N. L., & Islam, B. (2023) · Asia-Pacific Signal and Information Processing Association Annual Summit and Conference (APSIPA ASC) · doi:10.1109/apsipaasc58517.2023.10317201 · Cited by 10
“LIMU-BERT [32] is derived from bidirectional encoder representations from trans-formers (BERT) [33], and is used to pretrain transformers with unlabeled IMU data.”
MethodWei, Q., Huang, J., Gao, Y. C., & Dong, W. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3749509 · Cited by 21
A survey of foundation models for IoT: taxonomy and criteria-based analysis
Wei, H., Lee, D. Y., Rohal, S., Hu, Z., Rossi, R. A., Fang, S., & Pan, S. (2025) · CCF Transactions on Pervasive Computing and Interaction · doi:10.1007/s42486-025-00208-w · Cited by 13
Mouse2Vec: Learning Reusable Semantic Representations of Mouse Behaviour
Zhang, G., Hu, Z., Bâce, M., & Bulling, A. (2024) · CHI Conference on Human Factors in Computing Systems · doi:10.1145/3613904.3642141 · Cited by 13
Yang, B., Zhang, X., Zhang, J., et al. (2024) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3699768 · Cited by 5
Saha, S. S., Du, Y., Sandha, S. S., Garcia, L. A., Jawed, M. K., & Srivastava, M. B. (2023) · IEEE/ION Position Location and Navigation Symposium · doi:10.1109/plans53410.2023.10139997 · Cited by 13
“Possible solutions include the use of onboard BERT-like unsupervised pre-trained IMU embeddings [215], [216] and on-device training [181], [217]–[219].”
MethodGreta: Towards a General Roadside Unit Deployment Framework
Wu, X., Liu, Z., Li, Z., & Zhao, S. (2023) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2023.3335853 · Cited by 11
“Index Terms —Reinforcement learning, roadside unit deployment, vehicle-to-everything.”
BackgroundCHAR: Composite Head-Body Activities Recognition With a Single Earable Device
Zhu, P., Zhu, Y., Li, W., He, Y., Zou, Y., Wu, K., & Leung, V. C. M. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3548647 · Cited by 8
“LIMU-BERT [45] is a typical example of adopting self-supervised learning for human activity recognition.”
Hierarchical and Heterogeneous Federated Learning via a Learning-on-Model Paradigm
Shen, L., Yang, Q., Cui, K., Zheng, Y., Wei, X.-Y., Liu, J., & Han, J. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3581534 · Cited by 4
Physics-Informed Machine Learning Model Generalization in AIoT: Opportunites and Challenges
Tan, R., & Luo, W. (2023) · Cyber-Physical Systems and Internet of Things Week (CPS-IoT Week) · doi:10.1145/3576914.3588751 · Cited by 5
“Typically, self-supervised learning can reduce the requirement of the labeled training dataset by two orders of magnitude [3, 21].”
BackgroundInfluentialSelf-citationEnhancing the Applicability of Sign Language Translation
Li, J., Xu, J., Liu, Y., Xu, W., & Li, Z. (2024) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2024.3350111 · Cited by 6
“…analysis to detect potential privacy breaches when users enter sensitive information [24], [39], [40], [53], IoT foundation model construction [48], etc. Recent studies [25], [35], [36] have successfully reconstructed the arm’s movement trajectory using motion sensor data from a single…”
MultiHGR: Multi-Task Hand Gesture Recognition with Cross-Modal Wrist-Worn Devices
Lyu, M., Zhou, H., Guo, K., Zhou, W., Shen, X., & Gu, Y. (2024) · IEEE Conference on Computer Communications (INFOCOM) · doi:10.1109/infocom52122.2024.10621430 · Cited by 4
Towards Customizable Foundation Models for Human Activity Recognition with Wearable Devices
Qiu, M., Weng, C., Fan, M., & Wu, K. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3749479 · Cited by 4
“Additionally, HAR systems incorporate vision-based [32], wireless-based [16, 21, 50, 61, 66, 67], and mobile-based [17, 24, 27, 38, 40, 41, 43, 55, 57, 59] methods.”
InfluentialWang, X., Zhuang, Y., Huang, Y., Cao, X., Yu, T., Zhou, J., Zheng, Z., & El‐Sheimy, N. (2025) · IEEE Transactions on Aerospace and Electronic Systems · doi:10.1109/taes.2025.3550904 · Cited by 4
“This aspect of the MLM proves particularly beneficial for extracting crucial features from signals [51], which is the critical reason we chose BERT as the signal encoder.”
WiT: Wireless Tracking With Dual-Modality Transformers
Zhao, X., Wang, G., & Yang, L. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3636924 · Cited by 3
“Xu et al. [52] proposed LIMU-BERT, a model that leverages large volumes of unlabeled IMU data for pretraining, thus enhancing the robustness of localization systems while reducing the reliance on labeled data.”
Large Language Model-guided Semantic Alignment for Human Activity Recognition
Yan, H., Tan, H., Ding, Y. C., Zhou, P., Namboodiri, V., & Yang, Y. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3770652 · Cited by 31
“For data encoding, we employ a self-supervised learning model, BERT [7], as a representative method, following prior work [53] that masks and predicts parts of the sensor readings.”
InfluentialSelf-citationEarDA: Towards Accurate and Data-Efficient Earable Activity Sensing
Lyu, S., Chen, Y., Duan, D., Jia, R., & Xu, W. (2024) · doi:10.1109/cscaiot62585.2024.00005 · Cited by 15
“Thus, four datasets are chosen as they are widely utilized in previous works [14], [15].”
MethodMatching Skeleton-based Activity Representations with Heterogeneous Signals for HAR
Li, S., Zhang, J., Fu, X., Zhang, X., Shang, J., & Gupta, R. K. (2025) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3715014.3722089 · Cited by 5
SEGALL: A Unified Active Learning Framework for Wireless Sensing Data Segmentation
Zheng, N., Liu, R., Fan, X., Zhang, C., Zhang, L., & Yin, Z. (2025) · Proceedings of the ACM on Interactive Mobile Wearable and Ubiquitous Technologies · doi:10.1145/3749492 · Cited by 5
EasySpiro: Assessing Lung Function via Arbitrary Exhalations on Commodity Earphones
Xu, C., Xie, W., Yang, B., Zhang, Y., Gong, Y., Zhang, J., Li, W., Yang, S., & Zhang, Q. (2025) · ACM Annual International Conference on Mobile Computing and Networking (MobiCom) · doi:10.1145/3680207.3723489 · Cited by 7
“We pretrain this model on large IMU datasets [32, 44, 50, 53] using random masking as recommended in [69].”
TrackLet: Data-Driven Inertial Tracking on Your Own IMU Data
Wang, J., Zhao, Z., Cui, J., Lu, J., & Wu, B. Y. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3557190 · Cited by 2
From Vision to Motion: Translating Large-Scale Knowledge for Data-Scarce IMU Applications
Yoon, H., Cha, H., Nguyen, H. C., Gong, T., & Lee, S.-J. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3556998 · Cited by 10
Unveiling the Threat: Data-Free Backdoor Attacks on Pre-Trained Models for RF Fingerprinting
Zhao, T., Zhang, J., Dai, J., Sun, X., & Wang, X. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3628527 · Cited by 2
“The first term of the loss function ensures the backdoored PTM can generate benign representations for clean inputs, allowing the victim to accept it as the foundation model.”
Tanigaki, K., Maekawa, T., & Hara, T. (2026) · IEEE International Conference on Pervasive Computing and Communications · doi:10.1109/percom67906.2026.11524533 · Cited by 3
HARFMR: Human Activity Recognition with Feature Masking and Reconstruction
Cui, W., Chen, Y., Huang, Y. Q., Liu, C., & Zhu, T. (2024) · IFIP advances in information and communication technology · doi:10.1007/978-3-031-57919-6_6 · Cited by 2
Are You Being Tracked? Discover the Power of Zero-Shot Trajectory Tracing with LLMs!
Yang, H., Ji, S., Wu, R., & Xu, W. (2024) · doi:10.1109/cscaiot62585.2024.00008 · Cited by 15
The SES framework and Frequency domain information fusion strategy for Human activity recognition
Wu, H., Feng, H., Shi, L., Zhang, H., & Xu, H. (2024) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn60899.2024.10650417 · Cited by 2
“In recent work, [33] [34] have investigated efficient Transformer-related architectures in the HAR domain, proposing lightweight sensor Transformer architectures specifically designed for embedded mobile devices.”
BackgroundSelf-citationSensorGPT: A New Paradigm for Synthetic Sensory Data Generation
Sharma, R., Khamis, A., Khalifa, S., Bretherton, D., & Kusý, B. (2025) · IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) · doi:10.1109/percomworkshops65533.2025.00061 · Cited by 2
“Following the previous works, LIMU-BERT [29] and UniHAR [28], we evaluate the performance of SensorGPT on the three public datasets, UCI [20], HHAR [23], and MotionSense [16], in a human activity recognition (HAR) application.”
Comparing Self-Supervised Learning Techniques for Wearable Human Activity RecognitionPreprint
Ek, S., Presotto, R., Civitarese, G., Portet, F., Lalanda, P., & Bettini, C. (2023) · SSRN Electronic Journal · doi:10.2139/ssrn.4529608 · Cited by 2
CRFusion: Fine-Grained Object Identification Using RF-Image Modality Fusion
Xiao, L., Yang, Y., Chen, Z., Gao, Y., Mohapatra, P., & Hu, P. (2025) · IEEE Transactions on Mobile Computing · doi:10.1109/tmc.2025.3527872 · Cited by 5
“Furthermore, deep learning-based solutions can also be used to generate data samples or learn from unlabeled data [42], covering more combinations of object features while further reducing data collection costs.”
S2C‐HAR: A Semi‐Supervised Human Activity Recognition Framework Based on Contrastive Learning
Li, X., Nie, L., Liu, M., Cheng, W., Zhu, H., Wu, X., & Zhan, D. (2025) · Concurrency and Computation Practice and Experience · doi:10.1002/cpe.70027 · Cited by 3
“In the realm of HAR, data normalization like min-max normalization or mean-variance normalization [12] are often employed.”
InfluentialToward Foundation Models for Online Complex Event Detection in CPS-IoT: A Case Study
Han, L., Dong, G., Ouyang, X., Kaplan, L., Cerutti, F., & Srivastava, M. B. (2025) · doi:10.1145/3722565.3727198 · Cited by 4
“We use BEATs [2] and LIMU-Bert [23] to extract audio and IMU features, which are then processed through separate GRU layers, concatenated and fused in a fusion layer to form a 128-dimensional embedding 𝑣 𝑡 .”
Membership Inference Against Self-supervised IMU Sensing Applications
Zhao, T., Wang, N., & Wang, X. (2025) · ACM Conference on Embedded Networked Sensor Systems (SenSys) · doi:10.1145/3715014.3722060 · Cited by 4
Metadata
- Authors
- Xu, H., Zhou, P., Tan, R., Li, M., & Shen, G.
- Affiliations
- Nanyang Technological University (SG)
- Alibaba Group (China) (CN)
- Venue
- Proceedings of the 19th ACM Conference on Embedded Networked Sensor Systems
- Publisher
- ACM
- Published
- 2021-11-15
- Pages
- 220-233
- OpenAlex
- W3211771663
- Cited by
- 150
- References
- 56
- Citations by year
- 2026: 28 · 2025: 48 · 2024: 47 · 2023: 22 · 2022: 5
- FWCI
- 6.04
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- ACM Conference on Embedded Networked Sensor Systems (SenSys) (publishedVersion)
- Rare & Special e-Zone (The Hong Kong University of Science and Technology) (submittedVersion)
- Funders
- NTU CoE SUG
- National Research Foundation, Singapore under its Industry Alignment Fund ? Pre-positioning (IAF-PP) Funding Initiative
- Alibaba Group through Alibaba Innovative Research (AIR) Program and Alibaba-NTU Singapore Joint Research Institute (JRI)
- Topics
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Context-Aware Activity Recognition Systems
- Physical Sciences › Engineering › Aerospace Engineering › Inertial Sensor and Navigation
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Keywords
- inertial measurement unit, self-supervised learning, representation learning, mobile devices, human activity recognition, lightweight models, limited labeled data
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.