A5100448074- Advanced Image and Video Retrieval Techniques
- Advanced Vision and Imaging
- Domain Adaptation and Few-Shot Learning
- Face and Expression Recognition
- Multimodal Machine Learning Applications
- Machine Learning and ELM
- Robotics and Sensor-Based Localization
- Advanced Steganography and Watermarking Techniques
- Chaos-based Image/Signal Encryption
- Ammonia Synthesis and Nitrogen Reduction
- Visual Attention and Saliency Detection
- Remote-Sensing Image Classification
- 3D Shape Modeling and Analysis
- Video Surveillance and Tracking Methods
- Advanced Algorithms and Applications
- Digital Media Forensic Detection
- Advanced Photocatalysis Techniques
- Image and Video Quality Assessment
- Cancer-related molecular mechanisms research
- Multimedia Communication and Technology
- Hydrogen Storage and Materials
- Image and Video Stabilization
- Biomedical Research and Pathophysiology
- Epigenetics and DNA Methylation
- Image Processing Techniques and Applications
Anhui Medical University
2019-2025
China University of Mining and Technology
2010-2024
Shanghai University of Engineering Science
2023-2024
Ministry of Education of the People's Republic of China
2019-2023
Jiangsu Vocational Institute of Architectural Technology
2019-2023
Shanghai Jiao Tong University
2022
Obihiro University of Agriculture and Veterinary Medicine
2022
Shanghai First People's Hospital
2022
Hubei University of Arts and Science
2022
Wannan Medical College
2022
Abstract Acute kidney injury (AKI) is a destructive clinical condition induced by multiple insults including ischemic reperfusion, nephrotoxic drugs and sepsis. It characterized sudden decline in renal function, addition to excessive inflammation, oxidative stress programmed cell death of tubular epithelial cells. RIPK1-mediated necroptosis plays an important role AKI. In the present study, we evaluated treatment effects Compound-71 (Cpd-71), novel RIPK1 inhibitor, comparing with...
Traditional image steganography modifies the content of more or less, it is hard to resist detection steganalysis tools. To address this problem, a novel method named generative coverless information hiding based on adversarial networks proposed in paper. The main idea that class label replaced with secret as driver generate hidden directly, and then extract from through discriminator. It's first time achieved by networks. Compared traditional steganography, does not modify original image....
Diverse image captioning has achieved substantial progress in recent years. However, the discriminability of generative models and limitation cross entropy loss are generally overlooked traditional diverse models, which seriously hurts both diversity accuracy captioning. In this article, aiming to improve simultaneously, we propose a novel Conditional Variational Autoencoder (DCL-CVAE) framework for by seamlessly integrating sequential variational autoencoder with contrastive learning....
We propose a new method for modeling the indoor scene from single color image. With our system, user only needs to drag few semantic bounding boxes surrounding objects of interest. Our system then automatically finds most similar 3D models ShapeNet model repository and aligns them with corresponding To achieve this, each is represented as group view-dependent representations generated set synthesized views. iteratively conduct object segmentation retrieval, based on observation that good...
As is well known, traditional spectral clustering (SC) methods are developed based on the manifold assumption , namely, that two nearby data points in high-density region of a low-dimensional have same cluster label. But, for some high-dimensional and sparse data, such an might be invalid. Consequently, performance SC will degraded sharply this case. To solve problem, paper, we propose general embedded framework, which embeds true assignment matrix into nonlinear space by predefined...
We present in this paper an interactive approach for semantically modeling indoor environments given only a single image as input, without requiring access to the scene or using any additional measurements like RGBD cameras. Our key insight is that, although depth estimation from notoriously difficult, we can conveniently obtain relatively accurate normal map, which essentially conveys great deal of geometry. This enables us model each object data-driven manner by representing normal-based...
We propose a single image-based 3D model retrieval method for indoor scenes. By simulating the scene context of input image, our is able to handle several challenging scenarios featuring cluttered backgrounds and severe occlusions. To use system, user only needs drag few semantic bounding boxes query objects. The proposed approach then retrieves most similar models from ShapeNet repository, aligns them with corresponding objects automatically. This requires that are represented by calibrated...
As an important research field, face detection has been highly paid attentions by researchers. It theoretical value and application in computer vision pattern recognition technologies. Aimed at the problems over-training phenomenon, this paper presents improved sample training classifier, just considering feature uncertainty nearby threshold, these features corresponding samples adopted a new weight updating method. Experimental results show that classifier can obtain high rates than...
Transformer-based image captioning models have recently achieved remarkable performance by using new fully attentive paradigms. However, existing generally follow the conventional language model of predicting next word conditioned on visual features and partially generated words. They treat predictions nonvisual words equally usually tend to produce generic captions. To address these issues, we propose a novel part-of-speech-guided transformer (PoS-Transformer) framework for captioning....
Recently, transformer-based image captioning models have achieved significant performance improvement. However, due to the limitations of region visual features and deterministic projections between space caption space, existing methods still suffer from disentangled rigid sentences. To address these issues, we first introduce panoptic segmentation extract features, which can effectively alleviate confusion caused by widely-adopted features. Then, propose a based sequential conditional...
Dual-tone multi-frequency signal has been widely used in the modern communication systems. This paper gives a detailed analysis of characteristics dual-tone and method analyzing by software. Using Matlab this analyses signal, designs graphical interface imitating telephone using GUI toolbox, generates DTMF through sine wave superposition, introduces three frequency domain decoding approach, FFT, Goertzel Algorithm, improved Algorithm based on NDFT, results simulation algorithms, end,...