A5035137564- Advanced Sensor and Energy Harvesting Materials
- Gaze Tracking and Assistive Technology
- Tactile and Sensory Interactions
- Modular Robots and Swarm Intelligence
- Neuroscience and Neural Engineering
- Evolutionary Algorithms and Applications
- Organic Light-Emitting Diodes Research
- Insurance, Mortality, Demography, Risk Management
- Cancer-related molecular mechanisms research
- Legionella and Acanthamoeba research
- Conducting polymers and applications
- Machine Learning and Algorithms
- Privacy-Preserving Technologies in Data
- Game Theory and Voting Systems
- Genetics and Physical Performance
- Geoscience and Mining Technology
- Assistive Technology in Communication and Mobility
- Robotics and Automated Systems
- Granular flow and fluidized beds
- Renal and Vascular Pathologies
- Water Treatment and Disinfection
- Auction Theory and Applications
- Fecal contamination and water quality
- Machine Learning and Data Classification
- Civil and Geotechnical Engineering Research
Huazhong University of Science and Technology
2019-2024
Lanzhou Jiaotong University
2019
Research Center for Eco-Environmental Sciences
2016
Chinese Academy of Sciences
2016
Tsinghua University
2014
We demonstrate sizable, unencapsulated e-tattoos with transformable serpentine design and Cartan curve–inspired transfer.
Abstract The internal availability of silent speech serves as a translator for people with aphasia and keeps human–machine/human interactions working under various disturbances. This paper develops strategy to achieve all-weather, natural interactions. requires few usage specialized skills like sign language but accurately transfers high-capacity information in complicated changeable daily environments. In the strategy, tattoo-like electronics imperceptibly attached on facial skin record...
The facial expressions are a mirror of the elusive emotion hidden in mind, and thus, capturing is crucial way merging inward world virtual world. However, typical expression recognition (FER) systems restricted by environments where faces must be clearly seen for computer vision, or rigid devices that not suitable time-dynamic, curvilinear faces. Here, we present robust, highly wearable FER system based on deep-learning-assisted, soft epidermal electronics. electronics can fully conform...
Artificial intelligence of things systems equipped with flexible sensors can autonomously and intelligently detect the condition surroundings. However, current intelligent monitoring always rely on an external computer capability machine learning rather than integrating it into sensing device. The computer-assisted system is hampered by energy inefficiencies, privacy issues, bandwidth restrictions. Here, a flexible, large-scale array low-power in-sensor based compression hypervector encoder...
Abstract Six chest leads are the standardized clinical devices of diagnosing cardiac diseases. Emerging epidermal electronics technology shift dangling wires and bulky to imperceptible wearing, achieving both comfortable experience high-fidelity measuring. Extending small areas current scale requires eliminating interference from long interconnects rendering data acquisition (DAQ) portable. Herein, we developed a chest-scale electronic system (EES) for standard precordial-lead ECG hydration...
There is a relationship between biochemical function and chemical composition of corrosion scales, Fe<sub>3</sub>O<sub>4</sub>formation reduced iron release.
Federated learning (FL) allows agents to jointly train a global model without sharing their local data. However, due the heterogeneous nature of data, it is challenging optimize or even define fairness trained for agents. For instance, existing work usually considers accuracy equity as different in FL, which limited, especially under setting, since intuitively "unfair" enforce with high-quality data achieve similar those who contribute low-quality may discourage from participating FL. In...
Linear scalarization, i.e., combining all loss functions by a weighted sum, has been the default choice in literature of multi-task learning (MTL) since its inception. In recent years, there is surge interest developing Specialized Multi-Task Optimizers (SMTOs) that treat MTL as multi-objective optimization problem. However, it remains open whether fundamental advantage SMTOs over scalarization. fact, heated debates exist community comparing these two types algorithms, mostly from an...
Probabilistic circuits (PCs) have emerged as a powerful framework to compactly represent probability distributions for efficient and exact probabilistic inference. It has been shown that PCs with general directed acyclic graph (DAG) structure can be understood mixture of exponentially (in its height) many components, each which is product distribution over univariate marginals. However, existing learning algorithms often generate tree-structured or use intermediate steps compress them into...
We study the problem of allocating a set indivisible chores to three agents, among whom two have additive cost functions, in fair manner. Two fairness notions under consideration are envy-freeness up any chore (EFX) and relaxed notion, namely transferring (tEFX). In contrast case goods, remain relatively unexplored. particular, our results constructively prove existence tEFX allocation for agents if them functions ratio their highest lowest costs is bounded by two. addition, those identical...
To mitigate the bias exhibited by machine learning models, fairness criteria can be integrated into training process to ensure fair treatment across all demographics, but it often comes at expense of model performance. Understanding such tradeoffs, therefore, underlies design algorithms. this end, paper provides a complete characterization inherent tradeoff demographic parity on classification problems, under most general multi-group, multi-class, and noisy setting. Specifically, we show...
An integrated multifunctional epidermal electronics system for high-fidelity in situ personal health monitoring is introduced. It includes 3 kinds of sensors: the ECG sensor, temperature sensor and hydration sensor. can measure ECG, temperature, accurately.
Abstract With the rapid development of complexity and diversification robot services, large number data processing analysis problems occurred during working process. In this paper, we designed a collaborative system consisted multi-user multi-robot based on cloud platform. By pre-setting rights restrictions owned by each user in different scenarios, server determines scene which is located picture taken robot, thereby giving permissions. Enable to listen user’s instructions protect privacy....