arXiv subjects

Jianwei Tan

Publications and source records attributed to Jianwei Tan.

4 recordsLinked to original sources

arXiv2025

Differentially Private Recursive Least Squares Estimation for ARX Systems with Multi-Participants

This paper proposes a differentially private recursive least squares algorithm to estimate the parameter of autoregressive systems with exogenous inputs and multi-participants (MP-ARX systems) and protect each participant's sensitive information from potential attackers. We first give a rigorous differential privacy analysis of the algorithm, and establish the quantitative relationship between the added noises and the privacy-preserving level when the system is asymptotically stable. The asymptotic stability of the system is necessary for ensuring the differential privacy of the algorithm. We then give an estimation error analysis of the algorithm under the general and possible weakest excitation condition without requiring the boundedness, independence and stationarity on the regression vectors. Particularly, when there is no regression term in the system output and the differential privacy only on the system output is considered, $\varepsilon$-differential privacy and almost sure convergence of the algorithm can be established simultaneously. To minimize the estimation error of the algorithm with $\varepsilon$-differential privacy, the existence of the noise intensity is proved. Finally, two examples are given to show the efficiency of the algorithm.

eess.SY↗

arXiv2024

Subconscious Robotic Imitation Learning

Although robotic imitation learning (RIL) is promising for embodied intelligent robots, existing RIL approaches rely on computationally intensive multi-model trajectory predictions, resulting in slow execution and limited real-time responsiveness. Instead, human beings subconscious can constantly process and store vast amounts of information from their experiences, perceptions, and learning, allowing them to fulfill complex actions such as riding a bike, without consciously thinking about each. Inspired by this phenomenon in action neurology, we introduced subconscious robotic imitation learning (SRIL), wherein cognitive offloading was combined with historical action chunkings to reduce delays caused by model inferences, thereby accelerating task execution. This process was further enhanced by subconscious downsampling and pattern augmented learning policy wherein intent-rich information was addressed with quantized sampling techniques to improve manipulation efficiency. Experimental results demonstrated that execution speeds of the SRIL were 100\% to 200\% faster over SOTA policies for comprehensive dual-arm tasks, with consistently higher success rates.

cs.RO↗

arXiv2024

Constrained Behavior Cloning for Robotic Learning

Behavior cloning (BC) is a popular supervised imitation learning method in the societies of robotics, autonomous driving, etc., wherein complex skills can be learned by direct imitation from expert demonstrations. Despite its rapid development, it is still affected by limited field of view where accumulation of sensors and joint noise bring compounding errors. In this paper, we introduced geometrically and historically constrained behavior cloning (GHCBC) to dominantly consider high-level state information inspired by neuroscientists, wherein the geometrically constrained behavior cloning were used to geometrically constrain predicting poses, and the historically constrained behavior cloning were utilized to temporally constrain action sequences. The synergy between these two types of constrains enhanced the BC performance in terms of robustness and stability. Comprehensive experimental results showed that success rates were improved by 29.73% in simulation and 39.4% in real robot experiments in average, respectively, compared to state-of-the-art BC method, especially in long-term operational scenes, indicating great potential of using the GHCBC for robotic learning.

cs.RO↗

arXiv2024

ALow-Cost Real-Time Framework for Industrial Action Recognition Using Foundation Models

Action recognition (AR) in industrial environments -- particularly for identifying actions and operational gestures -- faces persistent challenges due to high deployment costs, poor cross-scenario generalization, and limited real-time performance. To address these issues, we propose a low-cost real-time framework for industrial action recognition using foundation models, denoted as LRIAR, to enhance recognition accuracy and transferability while minimizing human annotation and computational overhead. The proposed framework constructs an automatically labeled dataset by coupling Grounding DINO with the pretrained BLIP-2 image encoder, enabling efficient and scalable action labeling. Leveraging the constructed dataset, we train YOLOv5 for real-time action detection, and a Vision Transformer (ViT) classifier is deceloped via LoRA-based fine-tuning for action classification. Extensive experiments conducted in real-world industrial settings validate the effectiveness of LRIAR, demonstrating consistent improvements over state-of-the-art methods in recognition accuracy, scenario generalization, and deployment efficiency.

cs.CV↗