HELENA for 5G NR LEO NTN Channel Estimation: A Comparative Evaluation
研究测试HELENA在5G NR LEO NTN信道估计中的有效性，通过与其它模型对比，发现其在准确性和低延迟方面表现优异，无需针对NTN进行特定设计。
0 citationsRead paper
Selected work
Representative Papers
Predictive Coding-based Deep Neural Network Fine-tuning for Computationally Efficient Domain Adaptation
To address performance degradation of deep models under dynamic environmental shifts—such as sensor drift and illumination changes—this paper proposes a hybrid online domain adaptation method integrating backpropagation (BP) with predictive coding. The approach first establishes a robust representation via offline BP pretraining, then enables lightweight, local error-driven parameter updates through online predictive coding, balancing representational capacity and computational efficiency. Its key innovation lies in the first use of differentiable, low-overhead predictive coding as an online fine-tuning mechanism, specifically designed for resource-constrained edge devices and neuromorphic hardware. Experiments on MNIST and CIFAR-10 demonstrate that the method reduces computational cost by approximately 62% compared to pure BP-based online updating, while effectively mitigating accuracy loss and significantly enhancing model robustness and stability under continual distributional shift.
0 citationsRead paper
Towards Multimodal Social Conversations with Robots: Using Vision-Language Models
This study addresses the limitation of social robots in open-domain dialogue—overreliance on unimodal language models and insufficient visual perception and cross-modal understanding. We propose a multimodal social dialogue framework integrating large language models (LLMs) with vision-language models (VLMs), enabling real-time, synergistic comprehension and generation of visual and linguistic information through environment perception, cross-modal alignment, and dynamic context modeling. Our key contribution is the first systematic investigation of adaptive VLM deployment in general social interaction scenarios, explicitly identifying core technical requirements and challenges for multimodal social dialogue. Experimental results demonstrate significant improvements in dialogue naturalness, situational consistency, and user engagement. The framework establishes a scalable theoretical foundation and practical paradigm for embodied social intelligence.
0 citationsRead paper
Mixture of Autoencoder Experts Guidance using Unlabeled and Incomplete Data for Exploration in Reinforcement Learning
To address the challenges of sparse or absent rewards in reinforcement learning and unlabeled, incomplete expert demonstrations, this paper proposes the Mixture of Autoencoder Experts (MAE) model. MAE jointly models diverse behavioral patterns and missing information, mapping implicit state-wise similarity between agent trajectories and expert demonstrations onto intrinsic reward signals—enabling directed exploration without explicit labels or complete expert trajectories. By integrating autoencoder architectures with behavior similarity metrics, MAE efficiently encodes high-dimensional state-action spaces and generates shaped rewards. Experimental results demonstrate that MAE exhibits robust performance across both sparse- and dense-reward environments, significantly improving exploration efficiency and policy performance when expert demonstrations are scarce or of low quality.
0 citationsRead paper
Designing Robots with, not for: A Co-Design Framework for Empowering Interactions in Forensic Psychiatry
Forensic psychiatric inpatients frequently experience loss of control and chronic psychological stress within highly restrictive environments. Method: We propose a patient-empowerment-centered co-design framework, conducting three iterative participatory workshops with patients and clinical staff to develop speculative prototypes, enact contextual simulations, and integrate real-time affective feedback into the design cycle. Unlike top-down technological development, this approach navigates ethical and practical constraints inherent in high-risk clinical settings. Contribution/Results: The framework establishes functional requirements and responsive mechanisms for companion robots that balance autonomy, safety, and affective sensitivity. Empirical findings confirm patients’ capacity and willingness to meaningfully engage in technology design; resulting prototypes significantly improved intervention acceptability and human-centered alignment. This work contributes a reproducible, ethics-sensitive methodology for health robotics design in forensic mental healthcare.
0 citationsRead paper
Recent publications
Latest Papers
HELENA for 5G NR LEO NTN Channel Estimation: A Comparative Evaluation
研究测试HELENA在5G NR LEO NTN信道估计中的有效性，通过与其它模型对比，发现其在准确性和低延迟方面表现优异，无需针对NTN进行特定设计。
0 citationsRead paper
Predictive Coding-based Deep Neural Network Fine-tuning for Computationally Efficient Domain Adaptation
To address performance degradation of deep models under dynamic environmental shifts—such as sensor drift and illumination changes—this paper proposes a hybrid online domain adaptation method integrating backpropagation (BP) with predictive coding. The approach first establishes a robust representation via offline BP pretraining, then enables lightweight, local error-driven parameter updates through online predictive coding, balancing representational capacity and computational efficiency. Its key innovation lies in the first use of differentiable, low-overhead predictive coding as an online fine-tuning mechanism, specifically designed for resource-constrained edge devices and neuromorphic hardware. Experiments on MNIST and CIFAR-10 demonstrate that the method reduces computational cost by approximately 62% compared to pure BP-based online updating, while effectively mitigating accuracy loss and significantly enhancing model robustness and stability under continual distributional shift.
0 citationsRead paper
Towards Multimodal Social Conversations with Robots: Using Vision-Language Models
This study addresses the limitation of social robots in open-domain dialogue—overreliance on unimodal language models and insufficient visual perception and cross-modal understanding. We propose a multimodal social dialogue framework integrating large language models (LLMs) with vision-language models (VLMs), enabling real-time, synergistic comprehension and generation of visual and linguistic information through environment perception, cross-modal alignment, and dynamic context modeling. Our key contribution is the first systematic investigation of adaptive VLM deployment in general social interaction scenarios, explicitly identifying core technical requirements and challenges for multimodal social dialogue. Experimental results demonstrate significant improvements in dialogue naturalness, situational consistency, and user engagement. The framework establishes a scalable theoretical foundation and practical paradigm for embodied social intelligence.
0 citationsRead paper
Mixture of Autoencoder Experts Guidance using Unlabeled and Incomplete Data for Exploration in Reinforcement Learning
To address the challenges of sparse or absent rewards in reinforcement learning and unlabeled, incomplete expert demonstrations, this paper proposes the Mixture of Autoencoder Experts (MAE) model. MAE jointly models diverse behavioral patterns and missing information, mapping implicit state-wise similarity between agent trajectories and expert demonstrations onto intrinsic reward signals—enabling directed exploration without explicit labels or complete expert trajectories. By integrating autoencoder architectures with behavior similarity metrics, MAE efficiently encodes high-dimensional state-action spaces and generates shaped rewards. Experimental results demonstrate that MAE exhibits robust performance across both sparse- and dense-reward environments, significantly improving exploration efficiency and policy performance when expert demonstrations are scarce or of low quality.
0 citationsRead paper
Designing Robots with, not for: A Co-Design Framework for Empowering Interactions in Forensic Psychiatry
Forensic psychiatric inpatients frequently experience loss of control and chronic psychological stress within highly restrictive environments. Method: We propose a patient-empowerment-centered co-design framework, conducting three iterative participatory workshops with patients and clinical staff to develop speculative prototypes, enact contextual simulations, and integrate real-time affective feedback into the design cycle. Unlike top-down technological development, this approach navigates ethical and practical constraints inherent in high-risk clinical settings. Contribution/Results: The framework establishes functional requirements and responsive mechanisms for companion robots that balance autonomy, safety, and affective sensitivity. Empirical findings confirm patients’ capacity and willingness to meaningfully engage in technology design; resulting prototypes significantly improved intervention acceptability and human-centered alignment. This work contributes a reproducible, ethics-sensitive methodology for health robotics design in forensic mental healthcare.
0 citationsRead paper