Zhen Xing

ORCID: 0009-0008-0050-6323
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Human Pose and Action Recognition
  • Generative Adversarial Networks and Image Synthesis
  • Soil Moisture and Remote Sensing
  • Cancer-related molecular mechanisms research
  • Geoscience and Mining Technology
  • Myofascial pain diagnosis and treatment
  • Musculoskeletal pain and rehabilitation
  • Silicon Carbide Semiconductor Technologies
  • Advanced Sensor and Control Systems
  • Spine and Intervertebral Disc Pathology
  • Video Analysis and Summarization
  • RNA modifications and cancer
  • Lymphoma Diagnosis and Treatment
  • Gait Recognition and Analysis
  • Spectroscopy and Chemometric Analyses
  • Advanced Algorithms and Applications
  • Anomaly Detection Techniques and Applications
  • Multimodal Machine Learning Applications
  • Domain Adaptation and Few-Shot Learning
  • Robotics and Sensor-Based Localization
  • Cancer, Hypoxia, and Metabolism
  • Silicon and Solar Cell Technologies
  • Advanced Vision and Imaging
  • Advanced Data and IoT Technologies
  • Irrigation Practices and Water Management

Fudan University
2023-2025

Sun Yat-sen University
2025

First Affiliated Hospital of Fujian Medical University
2016-2024

Fujian Medical University
2016-2024

Beijing Academy of Agricultural and Forestry Sciences
2024

Ministry of Agriculture and Rural Affairs
2024

National Engineering Research Center for Information Technology in Agriculture
2016-2024

Chongqing University
2023-2024

Tiandi Science & Technology (China)
2023

China Coal Technology and Engineering Group Corp (China)
2023

Semi-supervised action recognition is a challenging but critical task due to the high cost of video annotations. Existing approaches mainly use convolutional neural networks, yet current revolutionary vision transformer models have been less explored. In this paper, we investigate under SSL setting for recognition. To end, introduce SVFormer, which adopts steady pseudo-labeling framework (i.e., EMA-Teacher) cope with unlabeled samples. While wide range data augmentations shown effective...

10.1109/cvpr52729.2023.01804 article EN 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2023-06-01

Pulmonary cryptococcosis (PC) was not a rare infectious disease in non-AIDS patients. However, data on the immune status were lacking southern China for comparative analysis of differences between immunocompromised and immunocompetent hosts. This study to investigate epidemiological, clinical, radiological, treatment profiles patients with PC.We performed retrospective review 88 diagnosed tissue-confirmed PC who HIV-infected from 2003 2013.Of patients, 35(39.7%) host. Fever CNS symptom...

10.21037/jtd.2016.10.36 article EN Journal of Thoracic Disease 2016-10-01

10.1109/icassp49660.2025.10890723 article EN ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) 2025-03-12

Diffusion models, as a type of generative model, have achieved impressive results in generating images and videos conditioned on textual conditions. However, the generation process diffusion models involves denoising dozens steps to produce photorealistic images/videos, which is computationally expensive. Unlike previous methods that design ``one-size-fits-all'' approaches for speed up, we argue should be sample-specific richness input texts. To this end, introduce AdaDiff, lightweight...

10.1609/aaai.v39i9.33075 article EN Proceedings of the AAAI Conference on Artificial Intelligence 2025-04-11

Abstract Band structure engineering by local strain of 2D transition metal dichalcogenides (TMDCs) is proven as an efficient means for improving the material performance in nanoscale light sources, photodetectors and flexible electronic devices. However, photoluminescence‐based techniques continue encountering challenges detecting changes band structures, mainly because indirect bandgap nature multilayer TMDCs well exciton diffusion at room temperature. Herein, it demonstrated that with help...

10.1002/adom.202500158 article EN Advanced Optical Materials 2025-05-12

(1) Background: The structure, function, and community interactions of soil microbial communities cultivated Meconopsis integrifolia were characterized by studying this alpine flower traditional endangered Tibetan medicine. (2) Methods: Soil bacteria fungi studied based on high-throughput sequencing technology. Bacteria isolated using culturomics functionally identified as IAA-producing, organic phosphorus-dissolving, inorganic iron-producing carriers. (3) Results: dominant bacterial phyla...

10.3390/biology12020160 article EN cc-by Biology 2023-01-19

In the seasonally frozen soil regions of northern China, silty clay is widely used as a subgrade bed filler in heavy-haul railway construction. this paper, influence freeze-thaw cycles on dynamic strength properties (strength parameters and critical stress) fillers before after cement improvement was investigated by series triaxial tests under different confining pressure conditions, test results were quantified to analyze effects improvement. The show that modification can significantly...

10.3390/su15043333 article EN Sustainability 2023-02-11

Purpose This bi-institutional study aimed to establish a robust model for predicting clinically significant prostate cancer (csPCa) (pathological grade group ≥ 2) in PI-RADS 3 lesions the transition zone by comparing performance of combination models. Materials and methods included 243 consecutive men who underwent 3-Tesla magnetic resonance imaging (MRI) ultrasound-guided transrectal biopsy from January 2020 April 2022 which is divided into training cohort 170 patients separate testing 73...

10.3389/fonc.2023.1247682 article EN cc-by Frontiers in Oncology 2023-11-23

Reconstructing detailed 3D objects from single-view images remains a challenging task due to the limited information available. In this paper, we introduce FDGaussian, novel two-stage framework for single-image reconstruction. Recent methods typically utilize pre-trained 2D diffusion models generate plausible views input image, yet they encounter issues with either multi-view inconsistency or lack of geometric fidelity. To overcome these challenges, propose an orthogonal plane decomposition...

10.48550/arxiv.2403.10242 preprint EN arXiv (Cornell University) 2024-03-15

The dynamic frost growth characteristics such as the thickness, mass on surface of fin-tube heat exchangers and refrigeration capacity were experimentally investigated different conditions air temperature, relative humidity, face velocity in psychrometric rooms. ethylene glycol–deionized water mixture (50% by volume) was used refrigerant a precise low temperature thermostat bath cold source. experimental results indicate that all pressure drop, decrease with but total increases which means...

10.1016/j.egypro.2017.03.1039 article EN Energy Procedia 2017-05-01

Text-guided video prediction (TVP) involves predicting the motion of future frames from initial frame according to an instruction, which has wide applications in virtual reality, robotics, and content creation. Previous TVP methods make significant breakthroughs by adapting Stable Diffusion for this task. However, they struggle with consistency temporal stability primarily due limited scale datasets. We observe that pretrained Image2Video diffusion models possess good priors dynamics but...

10.48550/arxiv.2406.06465 preprint EN arXiv (Cornell University) 2024-06-10

In order to obtain accurate real-time soil moisture data and the spatial distribution of moisture, measurement methods based on high-frequency capacitance edge field effect were analyzed, structure probe was studied, a multi-channel sensor designed. Moreover, with established two-dimensional trace planar model method finite element analysis, relationship between electric intensity studied amount most optimal determined. Design parameters which can achieve sensitivity detection range are:...

10.25165/ijabe.v9i3.2226 article EN International journal of agricultural and biological engineering 2016-05-31

Diffusion models have achieved significant success in image and video generation. This motivates a growing interest editing tasks, where videos are edited according to provided text descriptions. However, most existing approaches only focus on for short clips rely time-consuming tuning or inference. We the first propose Video Instruction (VIDiff), unified foundation model designed wide range of tasks. These tasks encompass both understanding (such as language-guided object segmentation)...

10.48550/arxiv.2311.18837 preprint EN other-oa arXiv (Cornell University) 2023-01-01

Multi-chips parallel IGBT module is widely used in new energy power grid connection, flexible AC and DC transmission, electric vehicles charging devices, rail transit traction other fields, the reliability of their operation very important. Therefore, how to select characteristic parameters that can characterize aging state multi-chips device particularly important for system active maintenance improvement. This paper takes as research object. According failure mode bond wire lift-off,...

10.1109/icpsasia58343.2023.10294789 article EN 2022 IEEE/IAS Industrial and Commercial Power System Asia (I&CPS Asia) 2023-07-07
Coming Soon ...