A5101834122- Advanced Image and Video Retrieval Techniques
- Topic Modeling
- Multimodal Machine Learning Applications
- Image and Signal Denoising Methods
- Photoacoustic and Ultrasonic Imaging
- Domain Adaptation and Few-Shot Learning
- Olfactory and Sensory Function Studies
- Digital Platforms and Economics
- Face and Expression Recognition
- Advanced Image Processing Techniques
- Visual Attention and Saliency Detection
- Nuclear Physics and Applications
- Advanced Bandit Algorithms Research
- Semiconductor materials and devices
- Urban Heat Island Mitigation
- Impact of Light on Environment and Health
- IoT-based Control Systems
- Blockchain Technology Applications and Security
- Surgical Simulation and Training
- Water Systems and Optimization
- Thermal Radiation and Cooling Technologies
- Human Motion and Animation
- Engine and Fuel Emissions
- FinTech, Crowdfunding, Digital Finance
- Supply Chain Resilience and Risk Management
Nanjing University
2023-2024
University College London
2023
Beijing University of Civil Engineering and Architecture
2023
Kindai University
2022
Tianjin University
2020-2021
Washington University in St. Louis
2021
Northwestern Polytechnical University
2021
University of Illinois Urbana-Champaign
2020
Dalian University of Technology
2013
Xi'an Railway Survey and Design Institute
2012
In this paper, we formulate saliency detection via absorbing Markov chain on an image graph model. We jointly consider the appearance divergence and spatial distribution of salient objects background. The virtual boundary nodes are chosen as in a absorbed time from each transient node to is computed. measures its global similarity with all nodes, thus can be consistently separated background when used metric. Since relies weights path their distance, region center may salient. further...
Daytime radiative cooling dissipates heat from surfaces by reflecting sunlight and emitting infrared radiation to outer space, featuring zero-energy consumption. Wood-based coolers have received more attention due their high emissivity, sustainability, low cost. However, they often degrade under ultraviolet (UV) exposure, resulting in a poor efficiency. Herein, inspired the structure-functionality relationship Saharan silver ants, an outdoor durable wood (DCW) is developed that achieves...
In this paper, we propose a bottom-up saliency model based on absorbing Markov chain (AMC). First, sparsely connected graph is constructed to capture the local context information of each node. All image boundary nodes and other are, respectively, treated as transient in chain. Then, expected number times from node all can be used represent value The absorbed time depends weights path their spatial coordinates, which are completely encoded transition probability matrix. Considering...
Although enterprises can promote dynamic capabilities by managing their innovation processes, the specific processes and ways to encourage such have been under-researched. This exploratory research illuminates relationship between through a case study of largest Chinese e-commerce platform, Taobao, vast complex digital platform with various actors interactions numerous related platforms. Based on previous research, interviews ten management personnel at Alibaba from April 2019 August 2021,...
Study Design. Prospective observational study. Objective. This article aims to develop a spatial augmented reality-based surgical navigation system assist in the placement of pedicle screws minimally invasive spine surgery and verify accuracy this method. Summary Background Data. Due their high good visualization ability, reality systems have been used surgeries. However, does not allow information be shared restricts doctors. Methods. A that implements based on projector can realize...
This work explores the zero-shot capabilities of foundation models in Visual Question Answering (VQA) tasks. We propose an adaptive multi-agent system, named Multi-Agent VQA, to overcome limitations object detection and counting by using specialized agents as tools. Unlike existing approaches, our study focuses on system's performance without fine-tuning it specific VQA datasets, making more practical robust open world. present preliminary experimental results under scenarios highlight some...
This study introduces a hypothesis-testing framework to assess whether large language models (LLMs) possess genuine reasoning abilities or primarily depend on token bias. We go beyond evaluating LLMs accuracy; rather, we aim investigate their bias in solving logical tasks. Specifically, develop carefully controlled synthetic datasets, featuring conjunction fallacy and syllogistic problems. Our outlines list of hypotheses where biases are readily identifiable, with all null assuming...
Context: Generative AI (GenAI) has emerged as a transformative tool in software engineering, with requirements engineering (RE) actively exploring its potential to revolutionize processes and outcomes. The integration of GenAI into RE presents both promising opportunities significant challenges that necessitate systematic analysis evaluation. Objective: This paper comprehensive literature review (SLR) analyzing state-of-the-art applications innovative proposals leveraging RE. It surveys...
This work presents an enhanced approach to generating scene graphs by incorporating a relationship hierarchy and commonsense knowledge. Specifically, we propose Bayesian classification head that exploits informative hierarchical structure. It jointly predicts the super-category or type of between two objects, along with detailed under each super-category. We design validation pipeline uses large language model critique results from graph prediction system then use feedback enhance...
Accurate anomaly detection of remote maintenance control system natural gas pipeline is great significance in ensuring the safe and stable operation network. A supervised method based on k-nearest neighbors searching clustering proposed. Firstly, labels neighbor samples are used to determine whether sample noise or belongs an cluster, iterative search conducted until no more abnormal belonging this cluster found. Then, filtered numbers new be generated each calculated. Finally, SMOTE...
When performing classification tasks, raw high dimensional features often contain redundant information, and lead to increased computational complexity overfitting. In this paper, we assume the data samples lie on a single underlying smooth manifold, define intra-class inter-class similarities using pairwise local kernel distances. We aim find linear projection maximize minimize simultaneously, so that projected low has optimized distances based label which is more suitable for Diffusion Map...
This paper presents a finding that leveraging the hierarchical structures among labels for relationships and objects can substantially improve performance of scene graph generation systems. The focus this work is to create an informative structure divide object relationship categories into disjoint super-categories in systematic way. Specifically, we introduce Bayesian prediction head jointly predict super-category between pair instances, as well detailed within simultaneously, facilitating...
Magnetic resonance imaging (MRI) has been applied in various fields, especially for the medical purposes. However, fine details and smooth areas of some critical patterns an MR image polluted by common thermal noise will interfere with diagnosis doctors. Thermal obeys Rician distribution, which is hard conventional denoising methods based on shift invariant spatial filtering approaches to dispose. Besides, detail edge information be inevitably damaged when smoothing noise, unacceptable...
This work presents an instance-agnostic learning framework that fuses vision with dynamics to simultaneously learn shape, pose trajectories, and physical properties via the use of geometry as a shared representation. Unlike many contact approaches assume motion capture input known shape prior for collision model, our proposed learns object's geometric dynamic from RGBD video, without requiring either category-level or instance-level priors. We integrate system, BundleSDF, ContactNets,...
This design aims to solve the problem of being unable move forward due obstacles during earthquakes. Through this program, unmanned vehicles can adjust their angles by identifying obstacles. article uses Arduino software for programming, achieving automatic recognition programs from different angles. Thus precise angle control and exploring differences in perspectives. Compared traditional manual driving, obstacle avoidance reduce risks accidents, thereby improving road safety. In event an...