Jun Wang

Jun Wang

Auto-generated profile, based on publicly available information

Jun Wang

School of Information Engineering, Nanchang Institute of Technology, Nanchang, China

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
5 April 2023 in IEEE Transactions on Multimedia
Verified
Journal Article
Closed access
JWJun Wang
Jun Wang
PYPeng Yin
Peng Yin
YWYuanyun Wang
Yuanyun Wang
WYWenhui Yang
Wenhui Yang
Convolutional Neural Networks (CNNs) and Transformer are two powerful representation learning techniques for visual tracking. Although CNNs can effectively reduce local redundancy via small-neighborhood convolution operations, their limited receptive fields make it difficult to capture global dependency. Self-attention in Transformer uses patches as the input representation, which can effectively capture long-range dependency. However, blind similarity comparisons between all patches can lead to high redundancy. Is there then a technique that combines well the advantages of both paradigms for visual tracking? In this work, we design a novel backbone network for feature extraction. First, we choose Depthwise Convolution and Pointwise Convolution to build a Convolution Mixer, which effectively separates spatial mixing from channel-wise mixing of information. The Convolution Mixer reduces redundancy in spatial and channel features while increasing receptive field. Then, to exploit the global modeling ability of self-attention, we construct a module by aggregating Convolution Mixer and self-attention. The module shares dominant computational complexity (the square of the channel size) in the first stage. In the second stage, the shift and summation operations are lightweight. Finally, to alleviate the overfitting of the backbone network during training, a dropout layer is added at the end of the module to improve the generalization ability of the network model. Stronger image features are provided for subsequent feature fusion and prediction. The proposed tracker (named CMAT) achieves satisfying tracking performance on ten challenging datasets. In particular, CMAT achieves a 64.1% AUC on LaSOT and a 68.9% AUC on UAV123 while running at 23 frames per second (FPS).

14 February 2023 in Knowledge-Based Systems
Verified
Journal Article
Closed access
YWYuanyun Wang
Yuanyun Wang
WZWenshuang Zhang
Wenshuang Zhang
CLChangwang Lai
Changwang Lai
JWJun Wang
Jun Wang

20 March 2024 in Expert Systems with Applications
Verified
Journal Article
Closed access
JWJun Wang
Jun Wang
SYShuai Yang
Shuai Yang
YWYuanyun Wang
Yuanyun Wang
GYGuang Yang
Guang Yang

7 April 2023 in Complex & Intelligent Systems
Verified
Journal Article
Open access
PDF file
JWJun Wang
Jun Wang
CLChangwang Lai
Changwang Lai
WZWenshuang Zhang
Wenshuang Zhang
YWYuanyun Wang
Yuanyun Wang
CMChenchen Meng
Chenchen Meng
Transformer-based trackers greatly improve tracking success rate and precision rate. Attention mechanism in Transformer can fully explore the context information across successive frames. Nevertheless, it ignores the equally important local information and structured spatial information. And irrelevant regions may also affect the template features and search region features. In this work, a multi-scale feature fusion network is designed with box attention and instance attention in Encoder–Decoder architecture based on Transformer. After extracting features, the local information and structured spatial information is learnt by multi-scale box attention, and the global context information is explored by instance attention. Box attention samples grid features from the region of interest. Therefore, it effectively focuses on the region of interest (ROI) and avoids the influence of irrelevant regions in feature extraction. At the same time, instance attention can also pay attention to the context information across successive frames, and avoid falling into local optimum. The long-range feature dependencies are learned in this stage. Extensive experiments are conducted on six challenging tracking datasets to demonstrate the superiority of the proposed tracker MDTT, including UAV123, GOT-10k, LaSOT, VOT2018, TrackingNet, and NfS. In particular, the proposed tracker achieves AUC score of $$64.7 \% $$ 64.7 % on LaSOT, $$78.1 \%$$ 78.1 % on TrackingNet and precision score of $$89.2 \%$$ 89.2 % on UAV123, which outperforms the baseline and most recent advanced trackers.

16 June 2017 in Intelligent Automation and Soft Computing
Verified
Journal Article
Closed access
JZJia Zhao
Jia Zhao
TFTanghuai Fan
Tanghuai Fan
LLLi Lü
Li Lü
HSHui Sun
Hui Sun
JWJun Wang
Jun Wang

10 June 2025 in IEEE Transactions on Circuits and Systems for Video Technology
Verified
Journal Article
Closed access
JWJun Wang
Jun Wang
BCBingfei Chai
Bingfei Chai
LZLingtao Zhou
Lingtao Zhou
YWYuanyun Wang
Yuanyun Wang
Despite the excellent performance achieved by Siamese tracking in most scenarios, its performance remains unsatisfactory in some scenarios. Traditional convolutional operations tend to use a fixed-size convolution kernel, which results in a small receptive field; thus, the model focuses only on local target features, keeping the model from considering global dependencies. Intercorrelation operations also do not explicitly model and capture information about local targets’ corner regions, which prevents the model from considering critical information about target edges and reduces the accuracy of locating the target. In this paper, we propose a feature extraction subnetwork based on a long-range spatial representation module that captures long-range spatial dependencies between the foreground and background. The network allows the model to learn more discriminative feature representations. We also construct a feature fusion network based on a local feature enhancement module that considers features contained in local targets’ corner regions more strongly. The proposed model can learn more comprehensive and detailed feature representations that lead to more accurate tracking. The proposed tracker is compared with SOTA trackers on six tracking datasets, and an average tracking speed of 45 FPS is achieved. Especially, it achieves 86.5% precision and 67.2% success rate on the UAV123 dataset with 40 FPS.