- Tactile and Sensory Interactions
- Interactive and Immersive Displays
- Gaze Tracking and Assistive Technology
- Augmented Reality Applications
- Hand Gesture Recognition Systems
- Human Pose and Action Recognition
- Innovative Human-Technology Interaction
- Anomaly Detection Techniques and Applications
- Caching and Content Delivery
- Human Motion and Animation
- Virtual Reality Applications and Impacts
- Peer-to-Peer Network Technologies
- Context-Aware Activity Recognition Systems
- Machine Fault Diagnosis Techniques
- Biometric Identification and Security
- Usability and User Interface Design
- Photoacoustic and Ultrasonic Imaging
- Spatial Cognition and Navigation
- Iterative Learning Control Systems
- Optical Coherence Tomography Applications
- Teleoperation and Haptic Systems
- Endodontics and Root Canal Treatments
- Digital Accessibility for Disabilities
- Advanced Image and Video Retrieval Techniques
- Advanced Graph Neural Networks
Shenzhen University Health Science Center
2025
Shanghai University
2025
Ningbo University
2025
Wenzhou Medical University
2023-2025
Hong Kong University of Science and Technology
2025
University of Hong Kong
2025
Tsinghua University
2006-2024
Soochow University
2022-2024
Beijing University of Posts and Telecommunications
2009-2023
Inner Mongolia University
2023
Popular online video-on-demand (VoD) services all maintain a large catalog of videos for their users to access. The knowledge video popularity is very important system operation , such as caching on content distribution network (CDN) servers. at given time quite well understood. We study how the changes with time, different types videos, and apply results design strategies. Our based analyzing access levels over data provided by service provider. main finding is, while there are variations,...
Abstract Background Homegardens are ecologically and culturally important systems for cultivating medicinal plants wellbeing by healers farmers in Naxi communities of the Sino Himalayan region. The cultivation associated ethnomedical knowledge base maintaining utilizing these resources is at risk with expanded commercialization natural resources, development policies rapid socio-economic change China. Research needed to understand plant species maintained homegardens, their use contribution...
We present DualRing, a novel ring-form input device that can capture the state and movement of user's hand fingers. With two IMU rings attached to thumb index finger, DualRing sense not only absolute gesture relative ground but also pose among segments. To enable natural thumb-to-finger interaction, we develop high-frequency AC circuit for on-body contact detection. Based on sensing information outline interaction space divide it into three sub-spaces: within-hand hand-to-surface...
We introduce UbiPhysio, a milestone framework that delivers fine-grained action description and feedback in natural language to support people's daily functioning, fitness, rehabilitation activities. This expert-like capability assists users properly executing actions maintaining engagement remote fitness programs. Specifically, the proposed UbiPhysio comprises descriptor knowledge retrieval-enhanced module. The translates data, represented by set of biomechanical movement features we...
We present HandSee, a novel sensing technique that can capture the state and movement of user's hands touching or gripping smartphone. place right angle prism mirror on front camera to achieve stereo vision scene above touchscreen surface. develop pipeline extract depth image from monocular RGB image, which consists three components: matching algorithm estimate pixel-wise scene, CNN-based online calibration detect hand skin, merging outputs hands. Building output, substantial set valuable...
Mid-air text entry on virtual keyboards suffers from the lack of tactile feedback, which brings challenges to both tap detection and input prediction. In this paper, we explored feasibility single-finger typing QWERTY in mid-air. We first conducted a study examine users' 3D behavior different sizes keyboards. Results showed that participants perceived vertical projection lowest point keyboard during as target location inferring taps based intersection between finger was not applicable....
Gestures performed accompanying the voice are essential for interaction to convey complementary semantics purposes such as wake-up state and input modality. In this paper, we investigated voice-accompanying hand-to-face (VAHF) gestures interaction. We targeted on because relate closely speech yield significant acoustic features (e.g., impeding propagation). conducted a user study explore design space of VAHF gestures, where first gathered candidate then applied structural analysis them in...
For iris recognition in non-cooperative environments, segmentation has been regarded as the first most important challenge still open to biometric community, affecting all downstream tasks from normalization recognition. In recent years, deep learning technologies have gained significant popularity among various computer vision and also introduced biometrics, especially segmentation. To investigate developments attract more interest of researchers method, we organized 2021 NIR Iris Challenge...
We present ShadowTouch, a novel sensing method to recognize the subtle hand-to-surface touch state for independent fingers based on optical auxiliary. ShadowTouch mounts forward-facing light source user's wrist construct shadows surface in front of when corresponding are close surface. With such an design, vertical movements near-surface magnified and turned shadow features cast surface, which recognizable computer vision algorithms. To efficiently each finger, we devised two-stage CNN-based...
In the past years, human computer interaction (HCI) research and innovation have developed substantially, leading to a number of fruitful topics. this paper, we surveyed HCI in China from 10-year perspective. We analyzed popular methodology topics among Chinese researchers, including modeling, user interface techniques, context awareness, acceptance performance, experience design, human-AI interaction, applications social influences. also conducted bibliography analysis on published papers...
Tracking fine-grained finger movements with IMUs for continuous 2D-cursor control poses significant challenges due to limited sensing capabilities. Our findings suggest that finger-motion patterns and the inherent structure of joints provide beneficial physical knowledge, which lead us enhance motion perception accuracy by integrating priors into ML models. We propose MouseRing, a novel ring-shaped IMU device enables finger-sliding on unmodified surfaces like touchpad. A dataset was created...
Large language models (LLMs) are widely adopted to generate synthetic datasets for various natural processing (NLP) tasks, such as text classification and summarization. However, accurately measuring the diversity of these datasets-an aspect crucial robust model performance-remains a significant challenge. In this paper, we introduce DCScore, novel method dataset from perspective. Specifically, DCScore formulates evaluation sample task, leveraging mutual relationships among samples. We...
Digital capturing of memorable personal items is a key way to archive memories. Although current digitization methods (e.g., photos, videos, 3D scanning) can replicate the physical appearance an item, they often cannot preserve its real-world interactivity. We present Interactive Item (IDI), concept reconstructing both and, more importantly, interactivity item. first conducted formative study understand users' expectations IDI, identifying features, including geometry, interfaces, and...
Large language models (LLMs) have demonstrated remarkable in-context reasoning capabilities across a wide range of tasks, particularly with unstructured inputs such as or images. However, LLMs struggle to handle structured data, graphs, due their lack understanding non-Euclidean structures. As result, without additional fine-tuning, performance significantly lags behind that graph neural networks (GNNs) in learning tasks. In this paper, we show on data can be conceptualized...
Digital twins in vehicle-in-the-loop (VIL) test has great practical significance for the functional development, testing and evaluation of intelligent vehicle. The study about credibility assessment dynamically evolving models still lacks effective approaches. In addition, it rarely been studied automotive tests. this paper, a closed loop dynamic virtual real-world interaction was built, its characteristics are also analyzed. According to methods, methodology based on information entropy is...
The National Medium and Long Term Education Reform Development Plan (2010-2020) puts forward the requirements of "renewing concept talent training deepening reform education system". Therefore, how to change promote nursing step by has become focus attention educators in recent years. goal results-oriented (OBE) is train students be specialized talents with strong comprehensive ability, which conforming needs today's society. Based on this, this paper analyzed current research trend guided...