Xiaoyun Zhang

Shanghai Jiao Tong University

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Zhiyong Gao

Shanghai Jiao Tong University

Ya Zhang

Shanghai Jiao Tong University

Li Chen

Wuhan University of Science and Technology

Yanfeng Wang

Shanghai Jiao Tong University

Chunlei Cai

Shanghai Jiao Tong University

Guo Lu

Shanghai Jiao Tong University

Yuncai Liu

Wenzhou Medical University

Wenbo Bao

Shanghai Jiao Tong University

Xuejiao Zhang

Institute of Applied Ecology

Siheng Chen

Shanghai Jiao Tong University

Cooperative Institutions

Shanghai Jiao Tong University

135

Nanjing Institute of Railway Technology

119

New Technology (Israel)

115

Chinese Academy of Sciences

Peking University

Zhejiang University

University of Chinese Academy of Sciences

Huazhong University of Science and Technology

Shandong Jiaotong University

Sichuan University

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Prediction of the maximum absorption wavelength of azobenzene dyes by QSPR tools

Spectrochimica Acta Part A Molecular and Biomolecular Spectroscopy (2011)

Xuan Xu Feng Luan Huitao Liu Jianbo Cheng Xiaoyun Zhang

Azobenzene

Loo

Molecular descriptor

10.1016/j.saa.2011.08.046

Cite

Citations (17)

Analysis and optimization of x265 encoder

Qiang Hu Xiaoyun Zhang Zhiyong Gao Jun Sun

x265 is an open-source encoder project which aims to deliver the world's fastest and most computationally efficient HEVC encoder. Although x265 has been developed efficiently with many optimization techniques, it is still not able to encode HD videos in real time even at its faster setting. In this paper, we deeply investigate the encoding framework and computational complexity of x265, and find that RDO process is the most time consuming part. Then, an efficient prediction scheme is proposed which includes decreasing the number of RDO times, early skip detection and fast intra mode decision. Experimental results show that the proposed method improves the speed of x265 from 19.86fps to 37.76fps for HD test sequences, i.e., 47.44% complexity reduction, with only 1.37% BDBR coding performance loss.

ENCODE

10.1109/vcip.2014.7051616

Cite

Citations (16)

Research progresses in the treatment of patella dislocation

Chin J Injury Repair and Wound Healing(Electronic Edition) (2018)

Zhiguang Zhang Shuxiong Bi Peng Hu Xiaoyun Zhang Shixiong Huo

As the increase of people′s participation in sports activities, the number of patients with patellar dislocation is continuously increasing and patellar dislocation most frequently presents in children and adolescence. Once patellar dislocation occurs, it will bring serious psychological and economic burden to the patients, and it may lead to osteoarthritis or chronic disability. Acute patellar dislocation after conservative treatment has a high recurrence rate. The major risk factors for the recurrence of dislocation include patellar dysplasia, patella alta, increased tibial tubercle-trochlear groove distance and patella tilt. The therapies of recurrent patellar dislocation include femoral troehleoplasty, tibial tubercle transfer, medical retinaculum placation, lateral retinaculum release and the medial patellofemoral ligament reconstruction, which is widly used in current clinical, etc. For patients with different lesions or pathological features of patellar dislocation, individualized treatment plans should be developed. Key words: Patellar dislocation; Knee joint; Surgical procedures, operative; MPFL

Medial Patellofemoral Ligament

Retinaculum

Patellofemoral joint

10.3877/cma.j.issn.1673-9450.2018.02.015

Cite

Citations (0)

Open-vocabulary Object Segmentation with Diffusion Models

arXiv (Cornell University) (2023)

Ziyi Li Qinye Zhou Xiaoyun Zhang Ya Zhang Yanfeng Wang

The goal of this paper is to extract the visual-language correspondence from a pre-trained text-to-image diffusion model, in the form of segmentation map, i.e., simultaneously generating images and segmentation masks for the corresponding visual entities described in the text prompt. We make the following contributions: (i) we pair the existing Stable Diffusion model with a novel grounding module, that can be trained to align the visual and textual embedding space of the diffusion model with only a small number of object categories; (ii) we establish an automatic pipeline for constructing a dataset, that consists of {image, segmentation mask, text prompt} triplets, to train the proposed grounding module; (iii) we evaluate the performance of open-vocabulary grounding on images generated from the text-to-image diffusion model and show that the module can well segment the objects of categories beyond seen ones at training time; (iv) we adopt the augmented diffusion model to build a synthetic semantic segmentation dataset, and show that, training a standard segmentation model on such dataset demonstrates competitive performance on the zero-shot segmentation(ZS3) benchmark, which opens up new opportunities for adopting the powerful diffusion model for discriminative tasks.

Discriminative model

Benchmark (surveying)

10.48550/arxiv.2301.05221

Cite

Citations (1)

In-loop perceptual model-based rate-distortion optimization for HEVC real-time encoder

Journal of Real-Time Image Processing (2018)

Qiang Hu Jun Zhou Xiaoyun Zhang Zhiyong Gao Ming–Ting Sun

Rate–distortion optimization

Algorithmic efficiency

10.1007/s11554-018-0772-1

Cite

Citations (6)

A Self-Powered Triboelectric Nanosensor for PH Detection

Journal of Nanomaterials (2016)

Ying Wu Yuanjie Su Junjie Bai Guang Zhu Xiaoyun Zhang

A self-powered, sliding electrification based triboelectric sensor was developed for detecting PH value from a periodic contact/separation motion. This innovative, cost-effective, simply designed sensor is composed of a fluorinated ethylene propylene thin film and an array of electrodes underneath. The operation of the TENG (triboelectric nanogenerator) sensor relies on a repetitive emerging-submerging process with traveling solution waves, in which the coupling between triboelectrification and electrostatic induction gives rise to alternating flows of electrons between electrodes. On the basis of coupling effect between triboelectrification and electrostatic induction, the sensor generates electric output signals which are associated with PH value. Experimental results show that the output voltage of the TENG sensor increases with the increasing PH value, which indicate that the PH value of different solution can be real-time monitored. This work not only demonstrates a new principle in the field of PH value measurement but also greatly expands the applicability of triboelectric nanogenerator (TENG) as self-powered sensors.

Nanogenerator

Contact electrification

Electrostatic induction

Nanosensor

Electromagnetic induction

10.1155/2016/5121572

Cite

Citations (53)

Motion estimation for depth maps coding based on rendered view distortion

International Conference on Audio, Language and Image Processing (2014)

Zhiru Shi Yunyu Shi Xiaoyun Zhang Zhiyong Gao

Incolour plus depth map 3D video, depth map is a luminance image that aids the virtual views generation. Since it is not the final image viewed by end user, the conventional depth map compression method, which minimizes the pixel errors of the depth map, introduces more distortions to the virtual rendered images. In this paper, a novel motion estimation scheme for depth map coding is proposed. By employing rendered view distortion in Rate-Distortion Optimization, the optimal Motion Vector is determined to improve the rendered views quality. An effective Lagrange multiplier is also derived. Experimental results illustrate that 34%-63% of bit can be saved comparing to the conventional motion estimation method.

Depth map

Motion vector

Distortion (music)

Quarter-pixel motion

Fast motion

Rate–distortion optimization

10.1109/icalip.2014.7009880

Cite

Citations (0)

Wood defect detection using wavelet and mathematical morphology

Computer Engineering and Applications Journal (2008)

Xiaoyun Zhang

Wood defect detection is a very important step in wood processing.In order to implement automatic detection,a method based on wavelet transform and mathematical morphology is presented in this paper.The method uses wavelet transform to suppress the interference information.Then bottom-hat transform of morphology and threshold processing are applied to get seeds in the wavelet reconstructed image.Eventually the wood defect can be segmented through region-growth algorithm.Experiments prove that this method can detect wood defect effectively and accurately and meet the requirement of wood processing.

Stationary wavelet transform

Second-generation wavelet transform

Wood processing

Continuous wavelet transform

Morphology

Harmonic wavelet transform

Source

Cite

Citations (0)

Task Decoupled Framework for Reference-based Super-Resolution

2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (2022)

Yixuan Huang Xiaoyun Zhang Yu Fu Siheng Chen Ya Zhang

Reference-based super-resolution(RefSR) has achieved impressive progress on the recovery of high-frequency details thanks to an additional reference high-resolution(HR) image input. Although the superiority compared with Single-Image Super-Resolution(SISR), existing RefSR methods easily result in the reference-underuse issue and the reference-misuse as shown in Fig. I. In this work, we deeply investigate the cause of the two issues and further propose a novel framework to mitigate them. Our studies find that the issues are mostly due to the improper coupled framework design of current methods. Those methods conduct the super-resolution task of the input low-resolution(LR) image and the texture transfer task from the reference image together in one module, easily introducing the interference between LR and reference features. Inspired by this finding, we propose a novel framework, which decouples the two tasks of RefSR, eliminating the interference between the LR image and the reference image. The super-resolution task upsamples the LR image leveraging only the LR image itself. The texture transfer task extracts and transfers abundant textures from the reference image to the coarsely upsampled result of the super-resolution task. Extensive experiments demonstrate clear improvements in both quantitative and qualitative evaluations over state-of-the-art methods.

10.1109/cvpr52688.2022.00584

Cite

Citations (18)

MEMC-Net: Motion Estimation and Motion Compensation Driven Neural Network for Video Interpolation and Enhancement

arXiv (Cornell University) (2018)

Wenbo Bao Wei‐Sheng Lai Xiaoyun Zhang Zhiyong Gao Ming–Hsuan Yang

Motion estimation (ME) and motion compensation (MC) have been widely used for classical video frame interpolation systems over the past decades. Recently, a number of data-driven frame interpolation methods based on convolutional neural networks have been proposed. However, existing learning based methods typically estimate either flow or compensation kernels, thereby limiting performance on both computational efficiency and interpolation accuracy. In this work, we propose a motion estimation and compensation driven neural network for video frame interpolation. A novel adaptive warping layer is developed to integrate both optical flow and interpolation kernels to synthesize target frame pixels. This layer is fully differentiable such that both the flow and kernel estimation networks can be optimized jointly. The proposed model benefits from the advantages of motion estimation and compensation methods without using hand-crafted features. Compared to existing methods, our approach is computationally efficient and able to generate more visually appealing results. Furthermore, the proposed MEMC-Net can be seamlessly adapted to several video enhancement tasks, e.g., super-resolution, denoising, and deblocking. Extensive quantitative and qualitative evaluations demonstrate that the proposed method performs favorably against the state-of-the-art video frame interpolation and enhancement algorithms on a wide range of datasets.

Interpolation

Motion interpolation

Quarter-pixel motion

Optical Flow

10.48550/arxiv.1810.08768

Cite

Citations (19)