Yimin Dai
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Yimin Dai has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
- Rui Tan
Nanyang Technological University
6 shared publications - Dongfang Guo
Nanyang technological university, Singapore, Singapore
2 shared publications - Pengfei Zhou
School of Computing and Information, University of Pittsburgh, Pittsburgh, Pennsylvania, USA
2 shared publications - Xin Lou
Singapore Institute of Technology, Singapore, Singapore
2 shared publications - Yuting Wu
Nanyang technological university, Singapore, Singapore
2 shared publications - Duc Phuc Nguyen
Ho Chi Minh City University of Technology (HCMUT), Ho Chi Minh City, Vietnam
1 shared publication
Latest Publications
- Open Access
- 20 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 108-109
Stable and Adaptive Edge Intelligence with Continuous-Time Neural Networks
- 20 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 108-109
Open AccessStable and Adaptive Edge Intelligence with Continuous-Time Neural Networks
- 5 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 2791-2804
A Continuous-Time On-Device Federated Learning Network
Time series is an important form of data generated by Internet-of-Things (IoT) devices. Closed-form continuous-time (CFC) neural networks offer superior expressivity for modeling time series data compared with recurrent neural networks. Additionally, their lower training and inference overhead make them well-suited for deployment on microcontroller-based IoT devices. This paper introduces FedCFC, an on-device federated learning network that operates based on the CFC models distributed across IoT devices. FedCFC incorporates a novel and communication-efficient aggregation strategy designed to mitigate the effects of class distribution imbalances across the participating IoT devices’ training data. The strategy is designed based on a new property of CFC identified in this paper, i.e., the insensitivity of a sub-model of CFC with respect to training data’s class distribution shift. Extensive evaluation with multiple time series datasets demonstrates that FedCFC attains comparable or superior accuracy while achieving a 7.6× to 11× reduction in communication overhead compared with recent federated learning approaches designed to address the class distribution skew problem. Furthermore, deployments of FedCFC on four IoT platforms highlight its suitability for resource-constrained devices with as little as 256 kB of memory or even less.
- 5 January 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Networking
- Vol. 34, 2791-2804
A Continuous-Time On-Device Federated Learning Network
Time series is an important form of data generated by Internet-of-Things (IoT) devices. Closed-form continuous-time (CFC) neural networks offer superior expressivity for modeling time series data compared with recurrent neural networks. Additionally, their lower training and inference overhead make them well-suited for deployment on microcontroller-based IoT devices. This paper introduces FedCFC, an on-device federated learning network that operates based on the CFC models distributed across IoT devices. FedCFC incorporates a novel and communication-efficient aggregation strategy designed to mitigate the effects of class distribution imbalances across the participating IoT devices’ training data. The strategy is designed based on a new property of CFC identified in this paper, i.e., the insensitivity of a sub-model of CFC with respect to training data’s class distribution shift. Extensive evaluation with multiple time series datasets demonstrates that FedCFC attains comparable or superior accuracy while achieving a 7.6× to 11× reduction in communication overhead compared with recent federated learning approaches designed to address the class distribution skew problem. Furthermore, deployments of FedCFC on four IoT platforms highlight its suitability for resource-constrained devices with as little as 256 kB of memory or even less.
- Open Access
- 3 September 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
- Vol. 9 (3) , 1-30
Certified Robustness against Sensor Heterogeneity in Acoustic Sensing
Domain shifts due to microphone hardware heterogeneity pose challenges to machine learning-based acoustic sensing. Existing methods enhance empirical performance but lack theoretical understanding. This paper proposes Certified Adaptive Physics-informed transform (CertiAPT), an approach that provides formal certification on the model accuracy and improves empirical performance against microphone-induced domain shifts. CertiAPT incorporates a novel Adaptive Physics-informed Transform (APT) to create transformations toward the target microphone without requiring application samples collected by the target microphone. It also establishes a theoretical upper bound on accuracy degradation due to microphone characteristic differences on unseen microphones. Furthermore, a robust training method with an APT gradient update scheme leverages APT and certification constraints to tighten the upper bound and improve empirical accuracy across sensor conditions. Extensive experiments on three acoustic sensing tasks, including keyword spotting, room recognition, and automated speech recognition, validate CertiAPT's certified robustness and show accuracy gains, compared with the latest approaches. Our implementation of CertiAPT is available at: https://github.com/bibom108/CertiAPT.
- 3 September 2025
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies
- Vol. 9 (3) , 1-30
Open AccessCertified Robustness against Sensor Heterogeneity in Acoustic Sensing
Domain shifts due to microphone hardware heterogeneity pose challenges to machine learning-based acoustic sensing. Existing methods enhance empirical performance but lack theoretical understanding. This paper proposes Certified Adaptive Physics-informed transform (CertiAPT), an approach that provides formal certification on the model accuracy and improves empirical performance against microphone-induced domain shifts. CertiAPT incorporates a novel Adaptive Physics-informed Transform (APT) to create transformations toward the target microphone without requiring application samples collected by the target microphone. It also establishes a theoretical upper bound on accuracy degradation due to microphone characteristic differences on unseen microphones. Furthermore, a robust training method with an APT gradient update scheme leverages APT and certification constraints to tighten the upper bound and improve empirical accuracy across sensor conditions. Extensive experiments on three acoustic sensing tasks, including keyword spotting, room recognition, and automated speech recognition, validate CertiAPT's certified robustness and show accuracy gains, compared with the latest approaches. Our implementation of CertiAPT is available at: https://github.com/bibom108/CertiAPT.
- Open Access
- 6 May 2025
- Published by Association for Computing Machinery (ACM)
- p. 345-356
Stochastic Differential Equation Networks for Time Series at Edge
- 6 May 2025
- Published by Association for Computing Machinery (ACM)
- p. 345-356
Open AccessStochastic Differential Equation Networks for Time Series at Edge
- Open Access
- 4 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 847-848
Demo: Invisible Adversarial Stripes against Traffic Sign Recognition in Autonomous Driving
Camera-based computer vision is crucial for autonomous vehicle perception. We demonstrate GhostStripe [5], an attack system that uses light-emitting diodes and exploits the camera's rolling shutter effect to generate adversarial stripes that are invisible to humans while misleading traffic sign recognition. To maintain stable attack effectiveness, GhostStripe controls the timing of the modulated light emission, adapting to both the camera's framing operation and the movement of the victim vehicle. Evaluated on real testbeds, GhostStripe can stably spoof traffic sign recognition results for up to 97% of frames to a wrong class when the victim vehicle passes the road section.
- 4 November 2024
- Published by Association for Computing Machinery (ACM)
- p. 847-848
Open AccessDemo: Invisible Adversarial Stripes against Traffic Sign Recognition in Autonomous Driving
Camera-based computer vision is crucial for autonomous vehicle perception. We demonstrate GhostStripe [5], an attack system that uses light-emitting diodes and exploits the camera's rolling shutter effect to generate adversarial stripes that are invisible to humans while misleading traffic sign recognition. To maintain stable attack effectiveness, GhostStripe controls the timing of the modulated light emission, adapting to both the camera's framing operation and the movement of the victim vehicle. Evaluated on real testbeds, GhostStripe can stably spoof traffic sign recognition results for up to 97% of frames to a wrong class when the victim vehicle passes the road section.
- Open Access
- 3 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 534-546
Invisible Optical Adversarial Stripes on Traffic Sign against Autonomous Vehicles
Camera-based computer vision is essential to autonomous vehicle's perception. This paper presents an attack that uses light-emitting diodes and exploits the camera's rolling shutter effect to create adversarial stripes in the captured images to mislead traffic sign recognition. The attack is stealthy because the stripes on the traffic sign are invisible to human. For the attack to be threatening, the recognition results need to be stable over consecutive image frames. To achieve this, we design and implement GhostStripe, an attack system that controls the timing of the modulated light emission to adapt to camera operations and victim vehicle movements. Evaluated on real testbeds, GhostStripe can stably spoof the traffic sign recognition results for up to 94% of frames to a wrong class when the victim vehicle passes the road section. In reality, such attack effect may fool victim vehicles into life-threatening incidents. We discuss the countermeasures at the levels of camera sensor, perception model, and autonomous driving system.
- 3 June 2024
- Published by Association for Computing Machinery (ACM)
- p. 534-546
Open AccessInvisible Optical Adversarial Stripes on Traffic Sign against Autonomous Vehicles
Camera-based computer vision is essential to autonomous vehicle's perception. This paper presents an attack that uses light-emitting diodes and exploits the camera's rolling shutter effect to create adversarial stripes in the captured images to mislead traffic sign recognition. The attack is stealthy because the stripes on the traffic sign are invisible to human. For the attack to be threatening, the recognition results need to be stable over consecutive image frames. To achieve this, we design and implement GhostStripe, an attack system that controls the timing of the modulated light emission to adapt to camera operations and victim vehicle movements. Evaluated on real testbeds, GhostStripe can stably spoof the traffic sign recognition results for up to 94% of frames to a wrong class when the victim vehicle passes the road section. In reality, such attack effect may fool victim vehicles into life-threatening incidents. We discuss the countermeasures at the levels of camera sensor, perception model, and autonomous driving system.
- Open Access
- 9 May 2023
- Published by Association for Computing Machinery (ACM)
- p. 123-135
Interpersonal Distance Tracking with mmWave Radar and IMUs
Tracking interpersonal distances is essential for real-time social distancing management and ex-post contact tracing to prevent spreads of contagious diseases. Bluetooth neighbor discovery has been employed for such purposes in combating COVID-19, but does not provide satisfactory spatiotemporal resolutions. This paper presents ImmTrack, a system that uses a millimeter wave radar and exploits the inertial measurement data from user-carried smartphones or wearables to track interpersonal distances. By matching the movement traces reconstructed from the radar and inertial data, the pseudo identities of the inertial data can be transferred to the radar sensing results in the global coordinate system. The re-identified, radar-sensed movement trajectories are then used to track interpersonal distances. In a broader sense, ImmTrack is the first system that fuses data from millimeter wave radar and inertial measurement units for simultaneous user tracking and re-identification. Evaluation with up to 27 people in various indoor/outdoor environments shows ImmTrack’s decimeters-seconds spatiotemporal accuracy in contact tracing, which is similar to that of the privacy-intrusive camera surveillance and significantly outperforms the Bluetooth neighbor discovery approach.
- 9 May 2023
- Published by Association for Computing Machinery (ACM)
- p. 123-135
Open AccessInterpersonal Distance Tracking with mmWave Radar and IMUs
Tracking interpersonal distances is essential for real-time social distancing management and ex-post contact tracing to prevent spreads of contagious diseases. Bluetooth neighbor discovery has been employed for such purposes in combating COVID-19, but does not provide satisfactory spatiotemporal resolutions. This paper presents ImmTrack, a system that uses a millimeter wave radar and exploits the inertial measurement data from user-carried smartphones or wearables to track interpersonal distances. By matching the movement traces reconstructed from the radar and inertial data, the pseudo identities of the inertial data can be transferred to the radar sensing results in the global coordinate system. The re-identified, radar-sensed movement trajectories are then used to track interpersonal distances. In a broader sense, ImmTrack is the first system that fuses data from millimeter wave radar and inertial measurement units for simultaneous user tracking and re-identification. Evaluation with up to 27 people in various indoor/outdoor environments shows ImmTrack’s decimeters-seconds spatiotemporal accuracy in contact tracing, which is similar to that of the privacy-intrusive camera surveillance and significantly outperforms the Bluetooth neighbor discovery approach.