HLHao Li

Hao Li

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Hao Li has a total of 30 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • CROP: Conservative Reward for Model-Based Offline Policy Optimization

    by Hao Li, Xiao-Hu Zhou, Shu-Hai Li, Mei-Jiang Gui, Xiao-Liang Xie, Shi-Qi Liu, Shuang-Yi Wang, Zhen-Qiu Feng, Zeng-Guang Hou

    Offline reinforcement learning (RL) aims to optimize a policy using collected data without online interactions. Model-based approaches are particularly appealing for addressing offline RL challenges because of their capability to mitigate the limitations of data coverage through data generation using models. Nonetheless, a prevalent issue in offline RL is the overestimation caused by distribution shift. This study proposes a novel model-based offline RL algorithm named Conservative Reward for model-based Offline Policy optimization (CROP). CROP introduces a streamlined objective that concurrently minimizes estimation error and the rewards of random actions, thereby yielding a robustly conservative reward estimator. Theoretical analysis shows that the designed conservative reward mechanism leads to a conservative policy evaluation and mitigates distribution shift. Experiments showcase that with the simple modification to reward estimation, CROP can conservatively estimate the reward and achieve competitive performance with existing methods. The source code will be available after acceptance.

  • Experience-Based Fuzzy Logic Framework for Robot-Assisted Endovascular Intervention

    by Pei-Liang Wu, Ding-Qian Wang, Xiao-Hu Zhou, Mei-Jiang Gui, Xiao-Liang Xie, Shi-Qi Liu, Shuang-Yi Wang, Hao Li, Zhi-Chao Lai, Zeng-Guang Hou

    Robot-assisted intervention at a higher level of autonomy can reduce the surgical complexity and physician workload. However, as the core requirement, the autonomous delivery of the guidewire remains challenging. The inherent flexibility of the guidewire complicates physical modeling, while existing learning-based approaches require prolonged training and exhibit limited interpretability. To address these issues, this paper proposes a novel method, experience-based Fuzzy LOgic framework for Robot-assisted endovascular Intervention (FLORI). FLORI emulates clinical operator techniques by leveraging a fuzzy logic system. It dynamically computes control parameters by analyzing both real-time guidewire tip positioning within the vascular architecture and historical attempt data. The proposed method maintains interpretability while improving the success rate and efficiency of guidewire delivery. Furthermore, this paper introduces a “macro + micro" human-machine collaborative delivery paradigm that allows physicians to switch between autonomous and manual delivery modes, as well as modify the guidewire delivery path during the surgical procedure. Physical experiments on a coronary artery model have demonstrated the effectiveness of FLORI and the role of the human-machine collaborative paradigm in reducing physician workload and enhancing the success rate of surgical procedures.


  • Upper Limb Motor Sequence Analysis: From Isolated to Sequential

    by Tian-Yu Xiang, Xiao-Hu Zhou, Mei-Jiang Gui, Xiao-Liang Xie, Shi-Qi Liu, Hao Li, De-Xing Huang, Jia-Xing Wang, Yong-Qiang Tang, Jia-Mou Liu, et al.

    Motor skills are performed through sequential movements rather than isolated actions. Yet, decoding these sequences from biosignals poses a significant challenge. To address this gap, this study transitions motor decoding from classifying movements in isolated time windows to segmenting sequential movements. The proposed algorithm segments the electromyography (EMG) sequence in a coarse-to-fine manner. It begins with frame-level segmentation and locating the approximate boundaries at the movement-level. A region-growing-inspired fusion strategy is then designed to incorporate the coarse segmentation and localization results for the fined output. Experiments on a self-collected EMG dataset demonstrate impressive results in segmenting movements for participant-dependent/independent setups (accuracy: $94.2{\%}/74.7{\%}$; dice coefficient: $92.5{\%}/61.7{\%}$; mean Intersection over Union: $80.9{\%}/51.9{\%}$). Further analysis shows the algorithm's ability to capture the natural rhythm in participants' movement sequences. This research paves the way for a deep understanding of motor sequences, which benefits various applications, such as rehabilitation engineering.


  • Learning Motor Cues in Brain-Muscle Modulation

    by Tian-Yu Xiang, Xiao-Hu Zhou, Xiao-Liang Xie, Shi-Qi Liu, Mei-Jiang Gui, Hao Li, De-Xing Huang, Xiu-Ling Liu, Zeng-Guang Hou

    Current studies for brain-muscle modulation often analyze selected properties in electrophysiological signals, leading to a partial understanding. This article proposes a cross-modal generative model that converts brain activities measured by electroencephalography (EEG) to corresponding muscular responses recorded by electromyography (EMG). Examining the generation process in the model highlights how the motor cue, representing implicit motor information hidden within brain activities, modulates the interaction between brain and muscle systems. The proposed model employs a two-stage generation process to bridge the semantic gap in cross-modal signals. Initially, the shared movement-related information between EEG and EMG signals is extracted using a contrastive learning framework. These shared representations act as conditional vectors in the subsequent EMG generation stage based on generative adversarial networks (GANs). Experiments on a self-collected multimodal electrophysiological signal data set show the algorithm’s superiority over existing time series generative methods in cross-modal EMG generation. Further insights derived from the model’s inference process underscore the brain’s strategy for muscle control during movements. This research provides a data-driven approach for the neuroscience community, offering a comprehensive perspective of brain-muscular modulation.


  • Learning EEG Motor Characteristics via Temporal-Spatial Representations

    by Tian-Yu Xiang, Xiao-Hu Zhou, Xiao-Liang Xie, Shi-Qi Liu, Hong-Jun Yang, Zhen-Qiu Feng, Mei-Jiang Gui, Hao Li, De-Xing Huang, Xiu-Ling Liu, et al.

    Electroencephalogram (EEG) is a widely used neural imaging technique for modeling motor characteristics. However, current studies have primarily focused on temporal representations of EEG, with less emphasis on the spatial and functional connections among electrodes. This study introduces a novel two-stream model to analyze both temporal and spatial representations of EEG for learning motor characteristics. Temporal representations are extracted with a set of convolutional neural networks (CNN) treated as dynamic filters, while spatial representations are learned by graph neural networks (GNN) using learnable adjacency matrices. At each stage, a res-block is designed to integrate temporal and spatial representations, facilitating a fusion of temporal-spatial information. Finally, the summarized representations of both streams are fused with fully connected neural networks to learn motor characteristics. Experimental evaluations on the Physionet, OpenBMI, and BCI Competition IV Dataset 2a demonstrate the model's efficacy, achieving accuracies of $73.6\%/70.4\%$ for four-class subject-dependent/independent paradigms, $84.2\%/82.0\%$ for two-class subject-dependent/independent paradigms, and 78.5% for a four-class subject-dependent paradigm, respectively. The encouraged results underscore the model's potential in understanding EEG-based motor characteristics, paving the way for advanced brain-computer interface systems.


  • Model-Based Offline Reinforcement Learning for Autonomous Delivery of Guidewire

    by Hao Li, Xiao-Hu Zhou, Xiao-Liang Xie, Shi-Qi Liu, Zhen-Qiu Feng, Mei-Jiang Gui, Tian-Yu Xiang, De-Xing Huang, Zeng-Guang Hou

    Guidewire delivery is a fundamental procedure in percutaneous coronary intervention. The inherent flexibility of the guidewire poses challenges in precise control, necessitating long-term training and substantial expertise. In response, this paper proposes a novel offline reinforcement learning (RL) algorithm, Conservative Offline Reinforcement Learning with Variational Environment Model (CORVE), for autonomous delivery of guidewire. CORVE first uses offline data to train an environment model and then optimizes the policy with both offline and model-generated data. The proposed method shares an encoder between the environmental model, policy, and Qfunction, mitigating the common sample inefficiency in imagebased RL. Besides, CORVE utilizes model prediction errors to forecast wrong deliveries in inference, which is an attribute absent in existing methods. The experimental results show that CORVE obtains superior performance in guidewire deliveries, achieving notably higher success rates and smoother movements than existing methods. These findings suggest that CORVE holds significant potential for enhancing the autonomy of vascular robotic systems in clinical settings.


  • Highly Interpretable Representation for Multi-Dimensional Tactile Perception

    by Mei-Jiang Gui, Xiao-Hu Zhou, Xiao-Liang Xie, Shi-Qi Liu, Zhen-Qiu Feng, Hao Li, Tian-Yu Xiang, De-Xing Huang, Bo-Xian Yao, Yong-Gen Ling, et al.

    Magnetic tactile perception systems have received increasing attention owing to their simple wiring framework and large-area sensing capabilities. However, existing systems often rely on data-driven methods, which is challenging to extract appropriate tactile representations, especially in complex interaction scenarios. To address such a challenge, this paper realizes a highly interpretable representation of the system’s two-stage conversion process (i.e., from changes in magnetic fields to spatial displacements and subsequently into tactile information) with the magnetic dipole model and dynamic Young’s modulus. Furthermore, the proposed representation method is incorporated into a novel spherical-array-based system for multi-dimensional tactile perception. Comprehensive experiments in simulated and real environments are conducted on four systems with various array arrangements. The proposed method can achieve relative errors of 0.54% and 1.75% under normal and tangential deformations, outperforming traditional data-driven approaches. It is envisaged that this study would benefit a wide range of industrial and domestic applications, such as remote surgery, dexterous manipulation, and human-robot interaction.


  • Open Access

    Learning Shared Semantic Information from Multimodal Bio-signals for Brain-Muscle Modulation Analysis

    by Tian-Yu Xiang, Xiao-Hu Zhou, Xiao-Liang Xie, Shi-Qi Liu, Hong-Jun Yang, Zhen-Qiu Feng, Mei-Jiang Gui, Hao Li, De-Xing Huang, Zeng-Guang Hou

    This paper presents a novel learning-based algorithm to investigate the high-level shared semantic information between electroencephalography (EEG) and electromyography (EMG) signals, for understanding brain-muscle modulation during movement execution. The proposed algorithm incorporates a spatial encoder that condenses spatial information obtained from EEG/EMG signals into unified temporal tokens using a learnable correlation matrix. These tokens are then encoded and decoded via a siamese temporal encoder and classification head to extract joint semantic information presented in cross-modal signals. Additionally, an analysis pipeline is designed to examine brain-muscle modulation based on the proposed algorithm. Experimental results from a self-collected multimodal bio-signals dataset validate the efficacy of the proposed algorithm in extracting and analyzing high-level latent semantic information shared in EEG and EMG signals, outperforming the state-of-the-art model by 5.35% in accuracy, 4.69% in precision, and 8.65% in recall. Notably, the designed analysis pipeline can also reveal low-level relationships, such as those related to time and space, between multimodal bio-signals. This research provides neuroscientists with a valuable tool for obtaining enhanced insights into brain-muscle modulation.


  • CASOG: Conservative Actor–Critic With SmOoth Gradient for Skill Learning in Robot-Assisted Intervention

    by Hao Li, Xiao-Hu Zhou, Xiao-Liang Xie, Shi-Qi Liu, Zhen-Qiu Feng, Zeng-Guang Hou

    The robot-assisted intervention has shown reduced radiation exposure to physicians and improved precision in clinical trials. However, existing vascular robotic systems follow master-slave control mode and entirely rely on manual commands. This article proposes a novel offline reinforcement learning algorithm, Conservative Actor–critic with SmOoth Gradient (CASOG), to learn manipulation skills on vascular robotic systems. The proposed algorithm conservatively estimates Q-function and smooths gradients of convolution layers to deal with distribution shift and overfitting issues. Furthermore, to focus on complex manipulations, transitions with larger absolute temporal-difference error are sampled with higher probability. Comparative experiments on multiple vascular models and offline data demonstrate that CASOG delivers guidewire to the target with higher success rates and fewer backward steps than prior offline reinforcement learning methods. These results indicate that the proposed algorithm is promising to improve the autonomy of vascular robotic systems.


  • Quantitative Movement Analysis Using Scaled Information Implied in Monocular Videos

    by Tian-Yu Xiang, Xiao-Hu Zhou, Xiao-Liang Xie, Shi-Qi Liu, Chen Wang, Mei-Jiang Gui, Hao Li, Zeng-Guang Hou

    Background: Quantitative movement analysis is significant in clinic decision-making for neurological diseases. However, current methods fail to be widely used due to expensive equipment and professional workers. Methods: To make movement analysis more accessible even in remote areas, this paper provides a quantitative movement analysis framework based monocular camera analyzing gait parameters automatically. A size-aligned method is formulated to recover scaled information uncertain in a single view under scaled restrictions from body size and geometric constraints among subject joints’ positions. The image key points are the subject joints’ positions in images and the scaling key points are processed by the size-aligned method. The attention mechanism fuses deep representations learned from image key points and scaling key points respectively. The fused representations are decoded to provide final analyses. Results: The suggested learning framework can analyze parameters with strong correlations to the ground truth according to experimental results on a dataset for cerebral palsy: cadence (r=0.81), speed (r=0.78), step length (r=0.78), knee flexion at maximum extension (r=0.85), gait deviation indexes (r=0.75), and asymmetric in gait deviation indexes (r=0.55). Conclusion: The encouraging findings show that it is feasible to quantitatively assess patients’ lower limb movement ability using a monocular camera.