Author name cluster

Lijun Guo

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

4 papers

1 author row

EAAI Journal 2026 Journal Article

Beyond reconstruction: Enhancing masked autoencoders with contrastive learning for video representation learning

Yawei Feng
Lijun Guo
Guitao Yu
Rong Zhang
Jiangbo Qian
Chong Wang
Shangce Gao

Self-supervised video representation learning primarily employs two methods: contrastive learning and masked video modeling, both of which possess unique advantages. Some studies have attempted to combine these two approaches to fully leverage their respective strengths. However, the intrinsic heterogeneity of these two methods poses challenges for existing models in integrating them, including complex model architectures, difficult training processes, and limited performance gains. To address these issues, this study proposes a novel video pre-training framework called Beyond Reconstruction (BR), which introduces a dual-track heterogeneous learning strategy. This strategy enables contrastive learning and masked video modeling to play their unique roles in different layers of Vision Transformers (ViTs), seamlessly integrating them into a unified framework to enhance the quality of video representations. Additionally, BR incorporates a motion-aware progressive masking strategy to strengthen spatiotemporal saliency modeling and stabilize the training process. By leveraging the advantages of contrastive learning in capturing global spatial motion objects, this strategy overcomes the limitations of previous masking methods. Experiments on multiple benchmarks, including action recognition and video object segmentation, show that the BR method achieves performance comparable to or even better than existing approaches under both fine-tuning and linear probing settings. These results demonstrate BR’s strong adaptability and efficiency in practical deployment: its stable fine-tuning performance enables effective adaptation to complex scenarios with limited annotations, while its strong linear probing capability allows the backbone to remain frozen, facilitating shared usage across multiple tasks and reducing overall computational cost without compromising performance.

Details DOI

EAAI Journal 2023 Journal Article

An adaptive position-guided gravitational search algorithm for function optimization and image threshold segmentation

Anjing Guo
Yirui Wang
Lijun Guo
Rong Zhang
Yang Yu
Shangce Gao

Gravitational search algorithm is a population-based optimization method. To address its low search performance and premature convergence, a novel variant called adaptive position-guided gravitational search algorithm is proposed. It utilizes the best, worst and other particles’ position information to adaptively determine the Kbest particles which provide a good movement direction. The gravitational force is reinforced by Kbest particles and new constructed Dbest particles to improve the exploration and exploitation abilities. Various particles’ position information jointly provide the effective search guideline and accelerate the convergence rate. Validations are conducted to firstly discuss the parameters and strategies of the proposed algorithm. Then, compared with several state-of-the-art gravitational search algorithm variants on CEC2017 benchmark functions, the proposed algorithm proves its superiority. Finally, the proposed algorithm exhibits the good segmentation effect on image threshold segmentation problems.

Details DOI

JBHI Journal 2023 Journal Article

VLTENet: A Deep-Learning-Based Vertebra Localization and Tilt Estimation Network for Automatic Cobb Angle Estimation

Lulin Zou
Lijun Guo
Rong Zhang
Lixin Ni
Zhenzuo Chen
Xiuchao He
Jianhua Wang

Scoliosis diagnosis and assessment rely upon Cobb angle estimation from X-ray images of the spine. Recently, automated scoliosis assessment has been greatly improved using deep learning methods. However, in such methods, the Cobb angle is usually predicted based on regression models that don't account for information of the spine structure. Alternatively, the Cobb angle can be estimated indirectly through landmark-detection and vertebra-segmentation, but this approach is still highly sensitive to small detection and segmentation errors. This paper proposes a novel deep-learning architecture, called the vertebra localization and tilt estimation network (VLTENet). This network boosts the Cobb angle estimation accuracy through employing vertebra localization and tilt estimation as network prediction goals. In particular, the VLTENet model innovatively combines a deep high-resolution network (HRNet) and a fully-convolutional U-Net architecture for capturing long-range contextual information, the overall structure, and local details in spinal X-ray images. A feature fusion channel attention (FFCA) module is also proposed to selectively emphasize more informative features and suppress less informative ones. In addition, a joint spine loss function (JS-Loss) is designed to account for the spine shape and other spatial constraints, so that the network focuses more on spine-related regions and ignore irrelevant background regions. Finally, we propose a new Cobb angle estimation method conforms with the clinical Cobb angle calculation guidelines, and produces accurate estimates for different types of scoliosis. Extensive experiments on the publically-available AASCE challenge dataset and on an in-house dataset demonstrated the superiority of our method for the task of automatic assessment of scoliosis.

Details DOI

JBHI Journal 2021 Journal Article

Discriminative Feature Network Based on a Hierarchical Attention Mechanism for Semantic Hippocampus Segmentation

Jiali Shi
Rong Zhang
Lijun Guo
Linlin Gao
Huifang Ma
Jianhua Wang

The morphological analysis of hippocampus is vital to various neurological studies including brain disorders and brain anatomy. To assist doctors in analyzing the shape and volume of the hippocampus, an accurate and automatic hippocampus segmentation method is highly demanded in the clinical practice. Given that fully convolutional networks (FCNs) have made significant contributions in biomedical image segmentation applications, we propose a notably discriminative feature network based on a hierarchical attention mechanism in hippocampal segmentation. First, considering the problem that the hippocampus is a rather small part in MR images, we design a context-aware high-level feature extraction module (CHFEM) to extract high-level features of scale invariance in the encoder stage. Further, we introduce a hierarchical attention mechanism into our segmentation framework. The mechanism is divided into three parts: a low-level feature spatial attention module (LFSAM) is developed to learn the spatial relationship between different pixels on each channel in the low-level stage of the encoder, a high-level feature channel attention module (HFCAM) is to model the semantic information relationship on different channel images in the high-level stage of the encoder, and a cross-connected attention module (CCAM) is designed in the decoder part to further suppress the noisy boundaries of hippocampus and simultaneously utilize the attentional low-level features from the encoder to better guide the high-level hippocampus edge segmentation in the decoder phase. The proposed approach achieves outstanding performance on the ADNI dataset and the Decathlon dataset compared with other semantic segmentation models and existing hippocampal segmentation approaches. Source code is available at https://github.com/LannyShi/Hippocampal-segmentation.

Details DOI