no code implementations • 11 Feb 2023 • Wei Chen, Chao Huang, Yanwei Yu, Yongguo Jiang, Junyu Dong
Trajectory-User Linking (TUL) is crucial for human mobility modeling by linking different trajectories to users with the exploration of complex mobility patterns.
no code implementations • 9 Jan 2023 • Meng Wang, Feng Gao, Junyu Dong, Heng-Chao Li, Qian Du
It is commonly nontrivial to build a robust self-supervised learning model for multisource data classification, due to the fact that the semantic similarities of neighborhood regions are not exploited in existing contrastive learning framework.
no code implementations • 29 Dec 2022 • Wei zhang, Yueyue Jiang, Junyu Dong, Xiaojiang Song, Renbo Pang, Boyu Guoan, Hui Yu
In this study, we developed the Multi-Task-Double Encoder Trajectory Gated Recurrent Unit (MT-DETrajGRU) model, which uses an improved double-encoder forecaster architecture to model the spatiotemporal sequence of the U and V components of the wind field; we designed a multi-task learning loss function to correct wind speed and wind direction simultaneously using only one model.
no code implementations • 16 Dec 2022 • Yakun Ju, Kin-Man Lam, Wuyuan Xie, Huiyu Zhou, Junyu Dong, Boxin Shi
We summarize the performance of deep learning photometric stereo models on the most widely-used benchmark data set.
no code implementations • 29 Sep 2022 • Xingshuai Dong, Matthew A. Garratt, Sreenatha G. Anavatti, Hussein A. Abbass, Junyu Dong
In order to aggregate the context information and edge attention features, we design a transformer-based feature aggregation module (TRFA).
1 code implementation • 12 Aug 2022 • Pengyang Yu, Chaofan Fu, Yanwei Yu, Chao Huang, Zhongying Zhao, Junyu Dong
Heterogeneous graph convolutional networks have gained great popularity in tackling various network analytical tasks on heterogeneous network data, ranging from link prediction to node classification.
1 code implementation • 9 Aug 2022 • Desen Meng, Feng Gao, Junyu Dong, Qian Du, Heng-Chao Li
To this end, we proposed a layer attention-based noise-tolerant network, termed LANTNet.
1 code implementation • 17 Jul 2022 • Haorui Song, Yong Du, Tianyi Xiang, Junyu Dong, Jing Qin, Shengfeng He
Consequently, in the decomposition phase, we further present a GAN prior based deghosting network for separating the final fine edited image from the coarse reconstruction.
1 code implementation • 8 May 2022 • Wei Chen, Shuzhe Li, Chao Huang, Yanwei Yu, Yongguo Jiang, Junyu Dong
In this paper, we propose a novel Mutual distillation learning network to solve the TUL problem for sparse check-in mobility data, named MainTUL.
no code implementations • 22 Apr 2022 • Xianglong, Yuezun Li, Haipeng Qu, Junyu Dong
However, the guidance map is fixed in existing methods, which can not consistently reflect the behavior of networks as the image is changed during iteration.
1 code implementation • 20 Apr 2022 • Zhongqiang Gao, Chuanqi Cheng, Yanwei Yu, Lei Cao, Chao Huang, Junyu Dong
We first categorize the temporal motifs based on their distinct properties, and then design customized algorithms that offer efficient strategies to exactly count the motif instances of each category.
no code implementations • 13 Mar 2022 • Junjie Wang, Feng Gao, Junyu Dong, Qian Du, Heng-Chao Li
We also propose the distinctive patch convolution for feature representation learning to reduce the time consumption.
no code implementations • 12 Mar 2022 • Lin Qi, Feng Gao, Junyu Dong, Xinbo Gao, Qian Du
Important findings on the use of spatial and spectral information in the autoencoder framework are discussed.
1 code implementation • 22 Jan 2022 • Junjie Wang, Feng Gao, Junyu Dong, Qian Du
Second, an adaptive DropBlock (AdapDrop) is proposed as a regularization method employed in the generator and discriminator to alleviate the mode collapse issue.
1 code implementation • 22 Jan 2022 • Yunhao Gao, Feng Gao, Junyu Dong, Heng-Chao Li
On the one hand, the multiscale capsule module is employed to exploit the spatial relationship of features.
1 code implementation • 22 Jan 2022 • Junjie Wang, Feng Gao, Junyu Dong, Shan Zhang, Qian Du
Synthetic aperture radar (SAR) image change detection is a vital yet challenging task in the field of remote sensing image analysis.
1 code implementation • 4 Nov 2021 • Yuxin Meng, Eric Rigall, Xueen Chen, Feng Gao, Junyu Dong, Sheng Chen
Physical modeling methods can offer the potential for extrapolation beyond observational conditions, while data-driven methods are flexible in adapting to data and are capable of detecting unexpected patterns.
1 code implementation • 18 Oct 2021 • Yunhao Gao, Feng Gao, Junyu Dong, Qian Du, Heng-Chao Li
Moreover, a correlation layer is designed to further explore the correlation between multitemporal images.
1 code implementation • 11 Oct 2021 • Fei Zhou, Xin Sun, Junyu Dong, Haoran Zhao, Xiao Xiang Zhu
Although Convolution Neural Networks (CNNs) has made substantial progress in the low-light image enhancement task, one critical problem of CNNs is the paradox of model complexity and performance.
no code implementations • 5 Aug 2021 • Cui Xie, Hao Guo, Junyu Dong
Ocean fronts can cause the accumulation of nutrients and affect the propagation of underwater sound, so high-precision ocean front detection is of great significance to the marine fishery and national defense fields.
no code implementations • 15 Jul 2021 • Yakun Ju, Muwei Jian, Shaoxiang Guo, YingYu Wang, Huiyu Zhou, Junyu Dong
In order to address this challenge, we here propose a photometric stereo network that incorporates Lambertian priors to better measure the surface normal.
no code implementations • 22 Jun 2021 • Ying Gao, Xiaohan Feng, Tiange Zhang, Eric Rigall, Huiyu Zhou, Lin Qi, Junyu Dong
Textures contain a wealth of image information and are widely used in various fields such as computer graphics and computer vision.
no code implementations • 21 Jun 2021 • Haoran Zhao, Xin Sun, Junyu Dong, Zihe Dong, Qiong Li
Recently, distillation approaches are suggested to extract general knowledge from a teacher network to guide a student network.
1 code implementation • 13 Jun 2021 • Yunhao Gao, Feng Gao, Junyu Dong, Heng-Chao Li
On the one hand, the capsule module is employed to exploit the spatial relationship of features.
1 code implementation • 11 Jun 2021 • Yanhai Gan, Xinghui Dong, Huiyu Zhou, Feng Gao, Junyu Dong
Based on this, we propose a general-purpose deep clustering framework which radically integrates representation learning and clustering into a single pipeline for the first time.
no code implementations • 27 Apr 2021 • Xin Sun, Zenghui Song, Yongbo Yu, Junyu Dong, Claudia Plant, Christian Boehm
This paper proposes a network embedding framework to capture the transfer behaviors on structured networks via deep prediction models.
1 code implementation • 19 Apr 2021 • Shengfeng He, Bing Peng, Junyu Dong, Yong Du
Shadow removal is an important yet challenging task in image processing and computer vision.
no code implementations • 18 Apr 2021 • Xin Sun, Changrui Chen, Xiaorui Wang, Junyu Dong, Huiyu Zhou, Sheng Chen
Furthermore, we also build a Gaussian dynamic pyramid Pooling to show its potential and generality in common semantic segmentation.
1 code implementation • 14 Apr 2021 • Xiaofan Qu, Feng Gao, Junyu Dong, Qian Du, Heng-Chao Li
In addition, we further propose a multi-region convolution module, which emphasizes the central region of each patch.
no code implementations • 12 Apr 2021 • Haoran Zhao, Xin Sun, Junyu Dong, Hui Yu, Huiyu Zhou
Then the generated samples are used to train the compact student network under the supervision of the teacher.
no code implementations • 6 Apr 2021 • Junjie Wang, Feng Gao, Junyu Dong
Convolutional neural networks (CNN) have made great progress for synthetic aperture radar (SAR) images change detection.
no code implementations • 6 Apr 2021 • Min Feng, Feng Gao, Jian Fang, Junyu Dong
An efficient linear self-attention fusion model is proposed in this paper for the task of hyperspectral image (HSI) and LiDAR data joint classification.
no code implementations • 6 Apr 2021 • Wenxia Liu, Feng Gao, Junyu Dong
As the ground objects become increasingly complex, the classification results obtained by single source remote sensing data can hardly meet the application requirements.
no code implementations • 6 Apr 2021 • Yan Gao, Feng Gao, Junyu Dong
Our network consists of the noise estimation subnetwork and denoising subnetwork.
1 code implementation • 29 Mar 2021 • Tiange Zhang, Feng Gao, Junyu Dong, Qian Du
Then, a new style discriminator is designed to improve the translation performance.
no code implementations • 18 Mar 2021 • Haoran Zhao, Kun Gong, Xin Sun, Junyu Dong, Hui Yu
The proposed approach promotes the performance of student model as the virtual sample created by multiple images produces a similar probability distribution in the teacher and student networks.
no code implementations • 20 Jan 2021 • Long Chen, Junyu Dong, Huiyu Zhou
CWSA is a new kind of data augmentation technique which augments the training data for the minority classes by generating various colors, textures and contrasts for the minority classes.
1 code implementation • 7 Jan 2021 • Aite Zhao, Jianbo Li, Junyu Dong, Lin Qi, Qianni Zhang, Ning li, Xin Wang, Huiyu Zhou
In recent years, single modality based gait recognition has been extensively explored in the analysis of medical images or other sensory data, and it is recognised that each of the established approaches has different strengths and weaknesses.
no code implementations • 7 Jan 2021 • Aite Zhao, Junyu Dong, Jianbo Li, Lin Qi, Huiyu Zhou
It is a challenging task to identify a person based on her/his gait patterns.
1 code implementation • ICCV 2021 • Zonghui Guo, Dongsheng Guo, Haiyong Zheng, Zhaorui Gu, Bing Zheng, Junyu Dong
Current solutions mainly adopt an encoder-decoder architecture with convolutional neural network (CNN) to capture the context of composite images, trying to understand what it looks like in the surrounding background near the foreground.
Ranked #5 on Image Harmonization on iHarmony4
1 code implementation • ICCV 2021 • Zhensheng Shi, Ju Liang, Qianqian Li, Haiyong Zheng, Zhaorui Gu, Junyu Dong, Bing Zheng
In this paper, we propose a novel multi-action relation model for videos, by leveraging both relational graph convolutional networks (GCNs) and video multi-modality.
no code implementations • ICCV 2021 • Zhanliang Wang, Junyu Dong, Xinguo Liu, Xueying Zeng
Our method is motivated by the recently proposed t-product based on any invertible linear transforms.
1 code implementation • 19 Oct 2020 • Long Chen, Feixiang Zhou, Shengke Wang, Junyu Dong, Ning li, Haiping Ma, Xin Wang, Huiyu Zhou
Moreover, inspired by the human education process that drives the learning from easy to hard concepts, we here propose the CMA training paradigm that first trains a clean detector which is free from the influence of noisy data.
1 code implementation • 2 Sep 2020 • Jianwen Lou, Xiaoxu Cai, Junyu Dong, Hui Yu
We propose to learn a cascade of globally-optimized modular boosted ferns (GoMBF) to solve multi-modal facial motion regression for real-time 3D facial tracking from a monocular RGB camera.
no code implementations • 2 Sep 2020 • Xiaoxu Cai, Hui Yu, Jianwen Lou, Xuguang Zhang, Gongfa Li, Junyu Dong
We present to recover the complete 3D facial geometry from a single depth view by proposing an Attention Guided Generative Adversarial Networks (AGGAN).
no code implementations • 21 Aug 2020 • Long Chen, Zheheng Jiang, Lei Tong, Zhihua Liu, Aite Zhao, Qianni Zhang, Junyu Dong, Huiyu Zhou
Underwater image enhancement, as a pre-processing step to improve the accuracy of the following object detection task, has drawn considerable attention in the field of underwater navigation and ocean exploration.
no code implementations • 29 Jun 2020 • Long Chen, Lei Tong, Feixiang Zhou, Zheheng Jiang, Zhenyang Li, Jialin Lv, Junyu Dong, Huiyu Zhou
To investigate how the underwater image enhancement methods influence the following underwater object detection tasks, in this paper, we provide a large-scale underwater object detection dataset with both bounding box annotations and high quality reference images, namely OUC dataset.
no code implementations • 18 Jun 2020 • Jinxuan Sun, Yang Chen, Junyu Dong, Guoqiang Zhong
Generative adversarial networks (GANs) are widely used in image generation tasks, yet the generated images are usually lack of texture details.
1 code implementation • 23 May 2020 • Long Chen, Zhihua Liu, Lei Tong, Zheheng Jiang, Shengke Wang, Junyu Dong, Huiyu Zhou
In addition, we propose a novel sample-weighted loss function which can model sample weights for SWIPENet, which uses a novel sample re-weighting algorithm, namely Invert Multi-Class Adaboost (IMA), to reduce the influence of noise on the proposed SWIPENet.
no code implementations • 18 Feb 2020 • Yu Zhang, Xin Sun, Junyu Dong, Changrui Chen, Yue Shen
The network first introduces a High-Order Representation module to extract the contextual high-order information from all stages of the backbone.
no code implementations • 13 Dec 2019 • Hongwei Xv, Xin Sun, Junyu Dong, Shu Zhang, Qiong Li
Low-shot learning indicates the ability to recognize unseen objects based on very limited labeled training samples, which simulates human visual intelligence.
no code implementations • 23 Jul 2019 • Xin Sun, Hongwei Xv, Junyu Dong, Qiong Li, Changrui Chen
Learning to recognize novel visual categories from a few examples is a challenging task for machines in real-world industrial applications.
1 code implementation • 23 Jul 2019 • Haoran Zhao, Xin Sun, Junyu Dong, Changrui Chen, Zihe Dong
Knowledge distillation aims to train a compact student network by transferring knowledge from a larger pre-trained teacher model.
no code implementations • 6 Jul 2018 • Yuanhong Xu, Pei Dong, Junyu Dong, Lin Qi
Obtaining dense 3D reconstrution with low computational cost is one of the important goals in the field of SLAM.
no code implementations • 19 May 2017 • Qin Zhang, Hui Wang, Junyu Dong, Guoqiang Zhong, Xin Sun
We formulate the SST prediction problem as a time series regression problem.
no code implementations • 24 Apr 2017 • Shu Zhang, Hui Yu, Ting Wang, Junyu Dong, Honghai Liu
With the increasing demands of applications in virtual reality such as 3D films, virtual Human-Machine Interactions and virtual agents, the analysis of 3D human face analysis is considered to be more and more important as a fundamental step for those virtual reality tasks.
no code implementations • 13 Apr 2017 • Junyu Dong, Li-Na Wang, Jun Liu, Xin Sun
Finally, given a set of semantic descriptions, the diverse properties of the samples in the semantic space can lead the framework to find an appropriate generation model that uses appropriate parameters to produce a desired texture.
no code implementations • 24 Mar 2017 • Yanhai Gan, Huifang Chi, Ying Gao, Jun Liu, Guoqiang Zhong, Junyu Dong
In this paper, we propose a joint deep network model that combines adversarial training and perceptual feature regression for texture generation, while only random noise and user-defined perceptual attributes are required as input.
no code implementations • 28 Feb 2017 • Long Chen, Junyu Dong, Shengke Wang, Kin-Man Lam, Muwei Jian, Hua Zhang, Xiaochun Cao
To bridge this gap, we introduce a cascaded structure to eliminate background and exploit a one-vs-rest loss to capture more minute variances among different subordinate categories.
no code implementations • 25 Nov 2016 • Guoqiang Zhong, Li-Na Wang, Junyu Dong
Since about 100 years ago, to learn the intrinsic structure of data, many representation learning approaches have been proposed, including both linear ones and nonlinear ones, supervised ones and unsupervised ones.
no code implementations • 22 Jul 2015 • Jianyuan Sun, Guoqiang Zhong, Junyu Dong, Yajuan Cai
Random forests are a type of ensemble method which makes predictions by combining the results of several independent trees.
no code implementations • 16 Jul 2015 • Guoqiang Zhong, Pan Yang, Sijiang Wang, Junyu Dong
For most existing hashing methods, an image is first encoded as a vector of hand-crafted visual feature, followed by a hash projection and quantization step to get the compact binary vector.
no code implementations • 14 May 2015 • Yanhai Gan, Jun Liu, Junyu Dong, Guoqiang Zhong
Particularly, each feature extraction stage includes two layers: a convolutional layer and a feature pooling layer.