Junfang Song

Dr. Junfang Song

orcid

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

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

Representative Publications

View All Publications
11 June 2025 in Applied Sciences
Verified
Journal Article
Open access
PDF file
JSJunfang Song
Junfang Song
JSJunfang Song
Junfang Song
FFFeiyang Feng
Feiyang Feng
ZYZhuyang Yan
Zhuyang Yan
AWAixin Wang
Aixin Wang
The rapid advancement in deep forgery technology in recent years has created highly deceptive face video content, posing significant security risks. Detecting these fakes is increasingly urgent and challenging. To improve the accuracy of deepfake face detection models and strengthen their resistance to adversarial attacks, this manuscript introduces a method for detecting forged faces and defending against adversarial attacks based on a multi-feature decision fusion. This approach allows for rapid detection of fake faces while effectively countering adversarial attacks. Firstly, an improved IMTCCN network was employed to precisely extract facial features, complemented by a diffusion model for noise reduction and artifact removal. Subsequently, the FG-TEFusionNet (Facial-geometry and Texture enhancement fusion-Net) model was developed for deepfake face detection and assessment. This model comprises two key modules: one for extracting temporal features between video frames and another for spatial features within frames. Initially, a facial geometry landmark calibration module based on the LRNet baseline framework ensured an accurate representation of facial geometry. A SENet attention mechanism was then integrated into the dual-stream RNN to enhance the model’s capability to extract inter-frame information and derive preliminary assessment results based on inter-frame relationships. Additionally, a Gram image texture feature module was designed and integrated into EfficientNet and the attention maps of WSDAN (Weakly Supervised Data Augmentation Network). This module aims to extract deep-level feature information from the texture structure of image frames, addressing the limitations of purely geometric features. The final decisions from both modules were integrated using a voting method, completing the deepfake face detection process. Ultimately, the model’s robustness was validated by generating adversarial samples using the I-FGSM algorithm and optimizing model performance through adversarial training. Extensive experiments demonstrated the superior performance and effectiveness of the proposed method across four subsets of FaceForensics++ and the Celeb-DF dataset.

8 April 2024 in Multimedia Systems
Verified
Journal Article
Closed access
PDF file
JSJunfang Song
Junfang Song
JSJunfang Song
Junfang Song
TWTengjiao Wang
Tengjiao Wang
FWFangxin Wang
Fangxin Wang
ZYZhuyang Yan
Zhuyang Yan

12 October 2024 in Electronics
Verified
Journal Article
Open access
PDF file
JSJunfang Song
Junfang Song
JSJunfang Song
Junfang Song
WWWenzhe Wu
Wenzhe Wu
Microexpressions are subtle facial movements that occur within an extremely brief time frame, often revealing suppressed emotions. These expressions hold significant importance across various fields, including security monitoring and human–computer interaction. However, the accuracy of microexpression recognition is severely constrained by the inherent characteristics of these expressions. To address the issue of low detection accuracy regarding the subtle features present in microexpressions’ facial action units, this paper proposes a microexpression action unit detection algorithm, Attention-embedded Dual Path and Shallow Three-stream Networks (ADP-DSTN), that incorporates an attention-embedded dual path and a shallow three-stream network. First, an attention mechanism was embedded after each Bottleneck layer in the foundational Dual Path Networks to extract static features representing subtle texture variations that have significant weights in the action units. Subsequently, a shallow three-stream 3D convolutional neural network was employed to extract optical flow features that were particularly sensitive to temporal and discriminative characteristics specific to microexpression action units. Finally, the acquired static facial feature vectors and optical flow feature vectors were concatenated to form a fused feature vector that encompassed more effective information for recognition. Each facial action unit was then trained individually to address the issue of weak correlations among the facial action units, thereby facilitating the classification of microexpression emotions. The experimental results demonstrated that the proposed method achieved great performance across several microexpression datasets. The unweighted average recall (UAR) values were 80.71%, 89.55%, 44.64%, 80.59%, and 88.32% for the SAMM, CASME II, CAS(ME)3, SMIC, and MEGC2019 datasets, respectively. The unweighted F1 scores (UF1) were 79.32%, 88.30%, 43.03%, 81.12%, and 88.95%, respectively. Furthermore, when compared to the benchmark model, our proposed model achieved better performance with lower computational complexity, characterized by a Floating Point Operations (FLOPs) value of 1087.350 M and a total of 6.356 × 106 model parameters.

1 July 2025 in AIP Advances
Verified
Journal Article
Open access
JSJunfang Song
Junfang Song
TWTengjiao Wang
Tengjiao Wang
JSJunfang Song
Junfang Song
ZYZhuyang Yan
Zhuyang Yan
3D reconstruction serves as a fundamental technology for comprehensively understanding and perceiving the physical world, with significant applications in computer vision fields such as industrial automation, autonomous navigation, and medical imaging. A critical challenge in monocular 3D reconstruction lies in the accurate and efficient estimation of camera intrinsic and extrinsic parameters, which are essential for determining precise positional information in each image. In this context, this paper proposes an improved PMVS (Patch-based Multi-view Stereo) 3D reconstruction assisted by union camera-turntable calibration, which can not only reduce manual effort and improve reconstruction efficiency but also capture rich surface details and accurate target localization, enabling the reconstruction of complex geometries. This study employs a cubic chessboard as a calibration object, replacing the traditional planar chessboard, which is positioned on a controllable turntable for calibration. This innovation allows for the determination of both intrinsic and extrinsic camera parameters from a single image while maintaining high accuracy. By integrating the rotary table, we can precisely control the inter-rotation angle to capture positional information and local texture from multiple perspectives. Furthermore, we utilize a transformer attention mechanism in place of conventional neural network (CNN) convolution to effectively capture global context and long-range dependencies within the images, thereby enhancing the quality of object surface reconstruction. Experimental results demonstrate that the proposed method significantly outperforms COLMAP and traditional PMVS 3D reconstruction techniques, yielding a 21.67% increase in the number of reconstructed point clouds and an ∼30% improvement in the structural similarity index (SSIM). In addition, this method reduces processing time by 24.67% while enhancing the visualization of complex surface reconstructions.