A5100600436- Multimodal Machine Learning Applications
- Advanced Image and Video Retrieval Techniques
- Domain Adaptation and Few-Shot Learning
- Natural Language Processing Techniques
- Topic Modeling
- Video Analysis and Summarization
- Photonic and Optical Devices
- Electrochemical Analysis and Applications
- Gyrotron and Vacuum Electronics Research
- Smart Grid Energy Management
- Photorefractive and Nonlinear Optics
- Particle accelerators and beam dynamics
- Image Retrieval and Classification Techniques
- Advanced Measurement and Detection Methods
- Power Line Inspection Robots
- Optical measurement and interference techniques
- Advanced MIMO Systems Optimization
- Image Enhancement Techniques
- Energy Harvesting in Wireless Networks
- Remote Sensing and Land Use
- Electrochemical sensors and biosensors
- Advanced Measurement and Metrology Techniques
- Protein purification and stability
- Recommender Systems and Techniques
- Modular Robots and Swarm Intelligence
Zhejiang University
2007-2024
Yunnan Forestry Vocational and Technical College
2024
Shanghai Jiao Tong University
2012-2018
Beijing Vaccum Electronic Research Institute
2013
Abstract Micro/nano robots have exhibited excellent application potential in the biomedical field, such as drug delivery, minimally invasive surgery, and bio-sensing. Furthermore, order to achieve practical application, it is essential for swimming micro/nano navigate towards specific targets or adjust their speed morphology complete environments. The navigation of with temporal spatial precision critical fulfilling demand applications. Here, we introduced a fully integrated wearable control...
Research on continual learning in multi-modal tasks has been receiving increasing attention. However, most existing work overlooks the explicit cross-modal and cross-task interactions. In this paper, we innovatively propose Low-rank Prompt Interaction (LPI) to address general problem of understanding, which considers both Specifically, as for former, employ correlation modules corresponding Transformer layers. Considering that training parameters scale number layers tasks, low-rank...
Recent methods for video question answering (VideoQA), aiming to generate answers based on given questions and content, have made significant progress in cross-modal interaction. From the perspective of understating, these existing frameworks concentrate various levels visual partially assisted by subtitles. However, audio information is also instrumental helping get correct answers, especially videos with real-life scenarios. Indeed, some cases, both contents are required complement each...
Transient, high repetition pulse laser can be applied to test numerous physical parameters, where in situ, real time measurement and isolation of vibration is highly demanded. Because its short half-width, power, repetition, even large distortion, the presents unique challenges conventional diagnosing methods. A system based on a novel cyclic radial shearing interferometer proposed diagnose transient, with common path, no reference plane, precision. With spatial-carrier methods, needs only...
CLIP-based models have made significant advancements in text-to-image retrieval tasks. However, these are typically trained on public datasets with optimizing all parameters, which limits their ability to generalize and adapt quickly personalized private datasets. In this paper, we introduce a lightweight federated learning solution, namely <u>Fed</u>erated <u>P</u>ersonalized <u>A</u>ugmentation <u>M</u>odel (<u>FedPAM</u>), achieve from multiple database. Specifically, for the query text,...
Generative retrieval has recently emerged as a promising approach to sequential recommendation, framing candidate item an autoregressive sequence generation problem. However, existing generative methods typically focus solely on either behavioral or semantic aspects of information, neglecting their complementary nature and thus resulting in limited effectiveness. To address this limitation, we introduce EAGER, novel recommendation framework that seamlessly integrates both information....
Abstract It has been well‐established that light‐matter interactions, as manifested by diverse linear and nonlinear optical (NLO) processes, are mediated real virtual particles, such electrons, phonons, excitons. Polarons, often regarded electrons dressed known to contribute exotic behaviors of solids, from superconductivity photocatalysis, while their role in materials’ NLO response remains largely unexplored. Here, the polarons supported a model ionic metal oxide, TiO 2 , is examined....
The integration of intelligent decision-making algorithms with urban cultural expression is becoming a hot topic in both academic and practical fields for exploring street landscapes. Exploring the application strategies regional landscape pattern recognition innovative design key step. single layout construction, deficiency, ecological imbalance, low resident participation seriously constrain overall quality improvement city. To address this dilemma, study delved into Kunming City selected...
Energy consumption in the base stations (BSs) recently has aroused significant concerns especially when renewable power been widely applied. In this paper, we jointly integrate from grid and energy to investigate management BSs with sleep- awake capability for cellular networks. our system model, are equipped two charging batteries operating at double timescales, exhibiting a more practical performance heterogeneous storage capability. We formulate problem as challenging nonlinear...
This paper focuses on intervention of collective behavior multi-agent systems. Distinct from the classical leader-follower method that guides group based inclination convergence, here by virtue collision avoidance in flocking behavior, we intervene adding some obstacles and designing their moving strategies. A model which can be treat as a shepherding process is developed, where sheep moves according to Boid model, shepherd aims drive all desired place. Two strategies with both global only...
Presently electricity utilities make use of a number inspection tools to survey their transmission lines and electrical distribution equipment. Visible ultraviolet cameras are the latest visual diagnostic available identify potential failures on Based DSP FPGA hardware platform, system is designed collect images from both visible channels. Moreover, it performs image fusion, characteristic enhancement analysis as well. For detecting corona, ultra-violet quantum detector used in form an...
Herein, a three-dimensional porous nanomaterials AuNPs-decorated polyaniline nanowire arrays/carbon nanotube (Au-PANI-array@CNT) was constructed by modified dilute polymerization and redox reaction. The Au-PANI-array@CNT exhibited larger specific surface area (50.03 m2 g-1) abundant mesopores (average pore size: 19.6 nm), due to the vertical growth of PANI arrays on CNT. composite with interconnected structure electroactive sites high adsorption capacity, significantly enhancing deposition...