Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Xiangyu Zhang

Xiangyu Zhang

The Threat of Offensive AI to Organizations


Jun 30, 2021
Yisroel Mirsky, Ambra Demontis, Jaidip Kotak, Ram Shankar, Deng Gelei, Liu Yang, Xiangyu Zhang, Wenke Lee, Yuval Elovici, Battista Biggio

Figure 1 for The Threat of Offensive AI to Organizations
Figure 2 for The Threat of Offensive AI to Organizations
Figure 3 for The Threat of Offensive AI to Organizations
Figure 4 for The Threat of Offensive AI to Organizations

  Access Paper or Ask Questions

PHYSFRAME: Type Checking Physical Frames of Reference for Robotic Systems


Jun 21, 2021
Sayali Kate, Michael Chinn, Hongjun Choi, Xiangyu Zhang, Sebastian Elbaum

Figure 1 for PHYSFRAME: Type Checking Physical Frames of Reference for Robotic Systems
Figure 2 for PHYSFRAME: Type Checking Physical Frames of Reference for Robotic Systems
Figure 3 for PHYSFRAME: Type Checking Physical Frames of Reference for Robotic Systems
Figure 4 for PHYSFRAME: Type Checking Physical Frames of Reference for Robotic Systems

  Access Paper or Ask Questions

SOLQ: Segmenting Objects by Learning Queries


Jun 09, 2021
Bin Dong, Fangao Zeng, Tiancai Wang, Xiangyu Zhang, Yichen Wei

Figure 1 for SOLQ: Segmenting Objects by Learning Queries
Figure 2 for SOLQ: Segmenting Objects by Learning Queries
Figure 3 for SOLQ: Segmenting Objects by Learning Queries
Figure 4 for SOLQ: Segmenting Objects by Learning Queries
* Tech Report.Code is available at https://github.com/megvii-research/SOLQ 
  Access Paper or Ask Questions

Fast Camera Image Denoising on Mobile GPUs with Deep Learning, Mobile AI 2021 Challenge: Report


May 17, 2021
Andrey Ignatov, Kim Byeoung-su, Radu Timofte, Angeline Pouget, Fenglong Song, Cheng Li, Shuai Xiao, Zhongqian Fu, Matteo Maggioni, Yibin Huang, Shen Cheng, Xin Lu, Yifeng Zhou, Liangyu Chen, Donghao Liu, Xiangyu Zhang, Haoqiang Fan, Jian Sun, Shuaicheng Liu, Minsu Kwon, Myungje Lee, Jaeyoon Yoo, Changbeom Kang, Shinjo Wang, Bin Huang, Tianbao Zhou, Shuai Liu, Lei Lei, Chaoyu Feng, Liguang Huang, Zhikun Lei, Feifei Chen

Figure 1 for Fast Camera Image Denoising on Mobile GPUs with Deep Learning, Mobile AI 2021 Challenge: Report
Figure 2 for Fast Camera Image Denoising on Mobile GPUs with Deep Learning, Mobile AI 2021 Challenge: Report
Figure 3 for Fast Camera Image Denoising on Mobile GPUs with Deep Learning, Mobile AI 2021 Challenge: Report
Figure 4 for Fast Camera Image Denoising on Mobile GPUs with Deep Learning, Mobile AI 2021 Challenge: Report
* Mobile AI 2021 Workshop and Challenges: https://ai-benchmark.com/workshops/mai/2021/. arXiv admin note: substantial text overlap with arXiv:2105.07809, arXiv:2105.07825 
  Access Paper or Ask Questions

MOTR: End-to-End Multiple-Object Tracking with TRansformer


May 07, 2021
Fangao Zeng, Bin Dong, Tiancai Wang, Cheng Chen, Xiangyu Zhang, Yichen Wei

Figure 1 for MOTR: End-to-End Multiple-Object Tracking with TRansformer
Figure 2 for MOTR: End-to-End Multiple-Object Tracking with TRansformer
Figure 3 for MOTR: End-to-End Multiple-Object Tracking with TRansformer
Figure 4 for MOTR: End-to-End Multiple-Object Tracking with TRansformer
* Tech Report. Code is available at https://github.com/megvii-model/MOTR 
  Access Paper or Ask Questions

RepMLP: Re-parameterizing Convolutions into Fully-connected Layers for Image Recognition


May 05, 2021
Xiaohan Ding, Xiangyu Zhang, Jungong Han, Guiguang Ding

Figure 1 for RepMLP: Re-parameterizing Convolutions into Fully-connected Layers for Image Recognition
Figure 2 for RepMLP: Re-parameterizing Convolutions into Fully-connected Layers for Image Recognition
Figure 3 for RepMLP: Re-parameterizing Convolutions into Fully-connected Layers for Image Recognition
Figure 4 for RepMLP: Re-parameterizing Convolutions into Fully-connected Layers for Image Recognition
* Work in progress 
  Access Paper or Ask Questions

Points as Queries: Weakly Semi-supervised Object Detection by Points


Apr 15, 2021
Liangyu Chen, Tong Yang, Xiangyu Zhang, Wei Zhang, Jian Sun

Figure 1 for Points as Queries: Weakly Semi-supervised Object Detection by Points
Figure 2 for Points as Queries: Weakly Semi-supervised Object Detection by Points
Figure 3 for Points as Queries: Weakly Semi-supervised Object Detection by Points
Figure 4 for Points as Queries: Weakly Semi-supervised Object Detection by Points

  Access Paper or Ask Questions

Diverse Branch Block: Building a Convolution as an Inception-like Unit


Mar 29, 2021
Xiaohan Ding, Xiangyu Zhang, Jungong Han, Guiguang Ding

Figure 1 for Diverse Branch Block: Building a Convolution as an Inception-like Unit
Figure 2 for Diverse Branch Block: Building a Convolution as an Inception-like Unit
Figure 3 for Diverse Branch Block: Building a Convolution as an Inception-like Unit
Figure 4 for Diverse Branch Block: Building a Convolution as an Inception-like Unit
* CVPR 2021 
  Access Paper or Ask Questions

Joint User Association and Power Allocation in Heterogeneous Ultra Dense Network via Semi-Supervised Representation Learning


Mar 29, 2021
Xiangyu Zhang, Zhengming Zhang, Luxi Yang

Figure 1 for Joint User Association and Power Allocation in Heterogeneous Ultra Dense Network via Semi-Supervised Representation Learning
Figure 2 for Joint User Association and Power Allocation in Heterogeneous Ultra Dense Network via Semi-Supervised Representation Learning
Figure 3 for Joint User Association and Power Allocation in Heterogeneous Ultra Dense Network via Semi-Supervised Representation Learning
Figure 4 for Joint User Association and Power Allocation in Heterogeneous Ultra Dense Network via Semi-Supervised Representation Learning

  Access Paper or Ask Questions

You Only Look One-level Feature


Mar 17, 2021
Qiang Chen, Yingming Wang, Tong Yang, Xiangyu Zhang, Jian Cheng, Jian Sun

Figure 1 for You Only Look One-level Feature
Figure 2 for You Only Look One-level Feature
Figure 3 for You Only Look One-level Feature
Figure 4 for You Only Look One-level Feature
* Accepted by CVPR2021. We provide more details and add additional comparison with YOLOv4 in this version 
  Access Paper or Ask Questions

EX-RAY: Distinguishing Injected Backdoor from Natural Features in Neural Networks by Examining Differential Feature Symmetry


Mar 17, 2021
Yingqi Liu, Guangyu Shen, Guanhong Tao, Zhenting Wang, Shiqing Ma, Xiangyu Zhang

Figure 1 for EX-RAY: Distinguishing Injected Backdoor from Natural Features in Neural Networks by Examining Differential Feature Symmetry
Figure 2 for EX-RAY: Distinguishing Injected Backdoor from Natural Features in Neural Networks by Examining Differential Feature Symmetry
Figure 3 for EX-RAY: Distinguishing Injected Backdoor from Natural Features in Neural Networks by Examining Differential Feature Symmetry
Figure 4 for EX-RAY: Distinguishing Injected Backdoor from Natural Features in Neural Networks by Examining Differential Feature Symmetry
* 21 pages 
  Access Paper or Ask Questions

Backdoor Scanning for Deep Neural Networks through K-Arm Optimization


Feb 09, 2021
Guangyu Shen, Yingqi Liu, Guanhong Tao, Shengwei An, Qiuling Xu, Siyuan Cheng, Shiqing Ma, Xiangyu Zhang

Figure 1 for Backdoor Scanning for Deep Neural Networks through K-Arm Optimization
Figure 2 for Backdoor Scanning for Deep Neural Networks through K-Arm Optimization
Figure 3 for Backdoor Scanning for Deep Neural Networks through K-Arm Optimization
Figure 4 for Backdoor Scanning for Deep Neural Networks through K-Arm Optimization

  Access Paper or Ask Questions

Neural Architecture Search with Random Labels


Jan 28, 2021
Xuanyang Zhang, Pengfei Hou, Xiangyu Zhang, Jian Sun

Figure 1 for Neural Architecture Search with Random Labels
Figure 2 for Neural Architecture Search with Random Labels
Figure 3 for Neural Architecture Search with Random Labels
Figure 4 for Neural Architecture Search with Random Labels
* Tech report 
  Access Paper or Ask Questions

RepVGG: Making VGG-style ConvNets Great Again


Jan 11, 2021
Xiaohan Ding, Xiangyu Zhang, Ningning Ma, Jungong Han, Guiguang Ding, Jian Sun

Figure 1 for RepVGG: Making VGG-style ConvNets Great Again
Figure 2 for RepVGG: Making VGG-style ConvNets Great Again
Figure 3 for RepVGG: Making VGG-style ConvNets Great Again
Figure 4 for RepVGG: Making VGG-style ConvNets Great Again
* The code and trained models are available at https://github.com/megvii-model/RepVGG 
  Access Paper or Ask Questions

Deep Feature Space Trojan Attack of Neural Networks by Controlled Detoxification


Jan 04, 2021
Siyuan Cheng, Yingqi Liu, Shiqing Ma, Xiangyu Zhang

Figure 1 for Deep Feature Space Trojan Attack of Neural Networks by Controlled Detoxification
Figure 2 for Deep Feature Space Trojan Attack of Neural Networks by Controlled Detoxification
Figure 3 for Deep Feature Space Trojan Attack of Neural Networks by Controlled Detoxification
Figure 4 for Deep Feature Space Trojan Attack of Neural Networks by Controlled Detoxification

  Access Paper or Ask Questions

Implicit Feature Pyramid Network for Object Detection


Dec 25, 2020
Tiancai Wang, Xiangyu Zhang, Jian Sun

Figure 1 for Implicit Feature Pyramid Network for Object Detection
Figure 2 for Implicit Feature Pyramid Network for Object Detection
Figure 3 for Implicit Feature Pyramid Network for Object Detection
Figure 4 for Implicit Feature Pyramid Network for Object Detection
* Tech report 
  Access Paper or Ask Questions

Rethinking Learnable Tree Filter for Generic Feature Transform


Dec 07, 2020
Lin Song, Yanwei Li, Zhengkai Jiang, Zeming Li, Xiangyu Zhang, Hongbin Sun, Jian Sun, Nanning Zheng

Figure 1 for Rethinking Learnable Tree Filter for Generic Feature Transform
Figure 2 for Rethinking Learnable Tree Filter for Generic Feature Transform
Figure 3 for Rethinking Learnable Tree Filter for Generic Feature Transform
Figure 4 for Rethinking Learnable Tree Filter for Generic Feature Transform
* Accepted by NeurIPS-2020 
  Access Paper or Ask Questions

Co-mining: Self-Supervised Learning for Sparsely Annotated Object Detection


Dec 03, 2020
Tiancai Wang, Tong Yang, Jiale Cao, Xiangyu Zhang

Figure 1 for Co-mining: Self-Supervised Learning for Sparsely Annotated Object Detection
Figure 2 for Co-mining: Self-Supervised Learning for Sparsely Annotated Object Detection
Figure 3 for Co-mining: Self-Supervised Learning for Sparsely Annotated Object Detection
Figure 4 for Co-mining: Self-Supervised Learning for Sparsely Annotated Object Detection
* Accepted to AAAI 2021 
  Access Paper or Ask Questions

Joint COCO and Mapillary Workshop at ICCV 2019: COCO Instance Segmentation Challenge Track


Oct 06, 2020
Zeming Li, Yuchen Ma, Yukang Chen, Xiangyu Zhang, Jian Sun

Figure 1 for Joint COCO and Mapillary Workshop at ICCV 2019: COCO Instance Segmentation Challenge Track
Figure 2 for Joint COCO and Mapillary Workshop at ICCV 2019: COCO Instance Segmentation Challenge Track
Figure 3 for Joint COCO and Mapillary Workshop at ICCV 2019: COCO Instance Segmentation Challenge Track
Figure 4 for Joint COCO and Mapillary Workshop at ICCV 2019: COCO Instance Segmentation Challenge Track
* 1st Place Technical Report in ICCV2019/ ECCV2020: MegDetV2 
  Access Paper or Ask Questions

EqCo: Equivalent Rules for Self-supervised Contrastive Learning


Oct 05, 2020
Benjin Zhu, Junqiang Huang, Zeming Li, Xiangyu Zhang, Jian Sun

Figure 1 for EqCo: Equivalent Rules for Self-supervised Contrastive Learning
Figure 2 for EqCo: Equivalent Rules for Self-supervised Contrastive Learning
Figure 3 for EqCo: Equivalent Rules for Self-supervised Contrastive Learning
Figure 4 for EqCo: Equivalent Rules for Self-supervised Contrastive Learning

  Access Paper or Ask Questions

MPG-Net: Multi-Prediction Guided Network for Segmentation of Retinal Layers in OCT Images


Sep 28, 2020
Zeyu Fu, Yang Sun, Xiangyu Zhang, Scott Stainton, Shaun Barney, Jeffry Hogg, William Innes, Satnam Dlay

Figure 1 for MPG-Net: Multi-Prediction Guided Network for Segmentation of Retinal Layers in OCT Images
Figure 2 for MPG-Net: Multi-Prediction Guided Network for Segmentation of Retinal Layers in OCT Images
Figure 3 for MPG-Net: Multi-Prediction Guided Network for Segmentation of Retinal Layers in OCT Images
* EUSIPCO2020 
  Access Paper or Ask Questions

Deep Learning & Software Engineering: State of Research and Future Directions


Sep 17, 2020
Prem Devanbu, Matthew Dwyer, Sebastian Elbaum, Michael Lowry, Kevin Moran, Denys Poshyvanyk, Baishakhi Ray, Rishabh Singh, Xiangyu Zhang

* Community Report from the 2019 NSF Workshop on Deep Learning & Software Engineering, 37 pages 
  Access Paper or Ask Questions

Activate or Not: Learning Customized Activation


Sep 10, 2020
Ningning Ma, Xiangyu Zhang, Jian Sun

Figure 1 for Activate or Not: Learning Customized Activation
Figure 2 for Activate or Not: Learning Customized Activation
Figure 3 for Activate or Not: Learning Customized Activation
Figure 4 for Activate or Not: Learning Customized Activation

  Access Paper or Ask Questions

Black-box Adversarial Sample Generation Based on Differential Evolution


Jul 30, 2020
Junyu Lin, Lei Xu, Yingqi Liu, Xiangyu Zhang

Figure 1 for Black-box Adversarial Sample Generation Based on Differential Evolution
Figure 2 for Black-box Adversarial Sample Generation Based on Differential Evolution
Figure 3 for Black-box Adversarial Sample Generation Based on Differential Evolution
Figure 4 for Black-box Adversarial Sample Generation Based on Differential Evolution
* 29 pages, 8 figures 
  Access Paper or Ask Questions

WeightNet: Revisiting the Design Space of Weight Networks


Jul 24, 2020
Ningning Ma, Xiangyu Zhang, Jiawei Huang, Jian Sun

Figure 1 for WeightNet: Revisiting the Design Space of Weight Networks
Figure 2 for WeightNet: Revisiting the Design Space of Weight Networks
Figure 3 for WeightNet: Revisiting the Design Space of Weight Networks
Figure 4 for WeightNet: Revisiting the Design Space of Weight Networks
* ECCV 2020 
  Access Paper or Ask Questions

Funnel Activation for Visual Recognition


Jul 24, 2020
Ningning Ma, Xiangyu Zhang, Jian Sun

Figure 1 for Funnel Activation for Visual Recognition
Figure 2 for Funnel Activation for Visual Recognition
Figure 3 for Funnel Activation for Visual Recognition
Figure 4 for Funnel Activation for Visual Recognition
* ECCV 2020 
  Access Paper or Ask Questions