Kai Wang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Kai Wang has a total of 42 co-authors. Here are the top co-authors by number of shared publications:
- Yongzhen Ke
Tiangong University, Tianjin, China
17 shared publications - Shuai Yang
School of Computer Science and Technology, Tiangong University, Tianjin, China
15 shared publications - Jing Guo
School of Computer Science and Technology, Tiangong University, Tianjin, China
4 shared publications - Fan Qin
Nankai University
3 shared publications - Hao Li
School of Computer Science and Technology, Tiangong University, Tiangong, China
2 shared publications - Jianghong Hu
Fitow (Tianjin) Detection Technology Co Ltd, Tianjin, China
2 shared publications
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 5389-5397
SGIN: Star-shaped Graph Interaction Network for Multi-attribute Aesthetic Quality Assessment of Portrait Sketches
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 5389-5397
Open AccessSGIN: Star-shaped Graph Interaction Network for Multi-attribute Aesthetic Quality Assessment of Portrait Sketches
- Open Access
- 1 July 2026
- Published by Elsevier in SSRN Electronic Journal
CFG-Net: A Coarse-to-Fine Grouping Network for Surface Defect Rating
Automatic surface defect rating is vital for industrial quality control but faces challenges in handling diverse defect scales and learning ordinal relationships with ambiguous boundaries. To address the inherent contradiction between scale and ordinality processing, we propose CFG-Net (Coarse-to-Fine Grouping Network), which adopts a "divide-and-conquer" strategy to decouple multi-scale rating tasks. CFG-Net uses a dual-stream grouping-gating module with lossless pixel unshuffle for initial-scale grouping and localization. Defects are dynamically dispatched to specialized branches: a low-level branch using anchor difference to amplify weak anomaly signals, and a high-level branch using deformable convolutions to fit irregular contours. Features are integrated via an adaptive ordinal fusion module guided by gating probabilities. Furthermore, we introduce a Multi-Margin N-Pair (MMNP) contrastive loss to construct an ordinal manifold space that strictly aligns with defect severity levels. Experimental results on real-world industrial datasets show that CFG-Net effectively resolves mixed-scale conflicts, achieving 97.2% group classification accuracy and 65.7% fine-grained rating accuracy. With a Mean Absolute Error (MAE) of 0.45 and Mean Squared Error (MSE) of 0.58, the proposed architecture demonstrates high precision and engineering robustness for complex, multi-scale defect rating tasks.
- 1 July 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessCFG-Net: A Coarse-to-Fine Grouping Network for Surface Defect Rating
Automatic surface defect rating is vital for industrial quality control but faces challenges in handling diverse defect scales and learning ordinal relationships with ambiguous boundaries. To address the inherent contradiction between scale and ordinality processing, we propose CFG-Net (Coarse-to-Fine Grouping Network), which adopts a "divide-and-conquer" strategy to decouple multi-scale rating tasks. CFG-Net uses a dual-stream grouping-gating module with lossless pixel unshuffle for initial-scale grouping and localization. Defects are dynamically dispatched to specialized branches: a low-level branch using anchor difference to amplify weak anomaly signals, and a high-level branch using deformable convolutions to fit irregular contours. Features are integrated via an adaptive ordinal fusion module guided by gating probabilities. Furthermore, we introduce a Multi-Margin N-Pair (MMNP) contrastive loss to construct an ordinal manifold space that strictly aligns with defect severity levels. Experimental results on real-world industrial datasets show that CFG-Net effectively resolves mixed-scale conflicts, achieving 97.2% group classification accuracy and 65.7% fine-grained rating accuracy. With a Mean Absolute Error (MAE) of 0.45 and Mean Squared Error (MSE) of 0.58, the proposed architecture demonstrates high precision and engineering robustness for complex, multi-scale defect rating tasks.
- 17 June 2026
- Published by World Scientific Pub Co Pte Ltd in International Journal of Software Engineering and Knowledge Engineering
- Vol. 37 (01) , 1-22
Group-Driven Causal Modeling and Root Cause Analysis for High-Dimensional Time Series
Root Cause Analysis (RCA) is vital for industrial monitoring and network diagnosis. In high-dimensional scenarios, traditional causal modeling is susceptible to noise and redundant dependencies, leading to unstable causal structures and poor interpretability. This paper proposes a novel method, GCRA (Group-Driven Causal Modeling and Root Cause Analysis), to address these challenges. First, the method utilizes cross-correlation to construct variable groups, establishing a group-level causal graph to effectively manage complex dependencies. Second, a mapping mechanism gradually refines this group-level graph into a variable-level structure. This structure guides the reconstruction of time-series observations, while structure-guided predictors are designed to model exogenous interventions representing unpredictable external disturbances. Finally, fine-grained RCA is achieved by calculating attribution scores based on these exogenous interventions. Experimental results across three synthetic and one real-world dataset demonstrate that GCRA accurately captures causal relationships and identifies root causes. The findings validate that group-level construction and structure-guided modeling significantly stabilize causal structures and enhance localization accuracy in high-dimensional data.
- 17 June 2026
- Published by World Scientific Pub Co Pte Ltd in International Journal of Software Engineering and Knowledge Engineering
- Vol. 37 (01) , 1-22
Group-Driven Causal Modeling and Root Cause Analysis for High-Dimensional Time Series
Root Cause Analysis (RCA) is vital for industrial monitoring and network diagnosis. In high-dimensional scenarios, traditional causal modeling is susceptible to noise and redundant dependencies, leading to unstable causal structures and poor interpretability. This paper proposes a novel method, GCRA (Group-Driven Causal Modeling and Root Cause Analysis), to address these challenges. First, the method utilizes cross-correlation to construct variable groups, establishing a group-level causal graph to effectively manage complex dependencies. Second, a mapping mechanism gradually refines this group-level graph into a variable-level structure. This structure guides the reconstruction of time-series observations, while structure-guided predictors are designed to model exogenous interventions representing unpredictable external disturbances. Finally, fine-grained RCA is achieved by calculating attribution scores based on these exogenous interventions. Experimental results across three synthetic and one real-world dataset demonstrate that GCRA accurately captures causal relationships and identifies root causes. The findings validate that group-level construction and structure-guided modeling significantly stabilize causal structures and enhance localization accuracy in high-dimensional data.
- 13 June 2026
- Published by Springer Nature in Applied Intelligence
- Vol. 56 (8) , 307
Reference image guided industrial defect generation with specified region and strength
- 13 June 2026
- Published by Springer Nature in Applied Intelligence
- Vol. 56 (8) , 307
Reference image guided industrial defect generation with specified region and strength
- 10 March 2026
- Published by Springer Nature in Multimedia Systems
- Vol. 32 (3) , 194
EFE-SDG: efficient feature extraction of finetuning-free model in subject-driven generation
- 10 March 2026
- Published by Springer Nature in Multimedia Systems
- Vol. 32 (3) , 194
EFE-SDG: efficient feature extraction of finetuning-free model in subject-driven generation
- 19 February 2026
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 20 (3) , 120
High-Fidelity Human Image Animation: Preserving Identity and Pose Consistency
- 19 February 2026
- Published by Springer Nature in Signal, Image and Video Processing
- Vol. 20 (3) , 120
High-Fidelity Human Image Animation: Preserving Identity and Pose Consistency
- 22 November 2025
- Published by Springer Nature in Pattern Analysis and Applications
- Vol. 28 (4) , 202
Stylized image generation based on multi-attribute decomposition
- 22 November 2025
- Published by Springer Nature in Pattern Analysis and Applications
- Vol. 28 (4) , 202
Stylized image generation based on multi-attribute decomposition
- Open Access
- 28 October 2025
- Published by Springer Nature in Research Square
EFE-SDG: Efficient Feature Extraction of Finetuning-Free Model in Subject-Driven Generation
Text-to-image models mark a major advance in AI, and subject-driven generation techniques hold promise. The subject-driven models that are currently avail- able still necessitate a compromise between the fidelity of the subject and the extent of text-to-image generation. Furthermore, they cannot provide a compre- hensive understanding of the reference image based on a limited amount of data. To address these limitations, we propose the EFE-SDG model, which is divided into three blocks: (i) the MOCI block employs the Gpt-4o and other large mod- els for the processing of the reference image to obtain more information about the reference image. (ii) The ASCA module, based on the decoupling strategy of the cross-attention module, performs additional processing on the reference image to obtain more detailed and richer high-level features of the subject. (iii) The subject Feature Adaptive Attention Rules is used to fuse the low-level fea- tures extracted by the ref-diffusion model and the high-level features extracted by ASCA at the Attention layer of the main-diffusion model. In addition, we use the ref-diffusion model to extract low-level feature inputs from the reference features to the main-diffusion model, which circumvents the different training distributions of the other encoders and the main-diffusion model. In our compar- ative experiments, with respect to graphic alignment, our approach demonstrates performance similar to that of the other methods, but our approach uses sub- stantially smaller datasets and fewer computational resources during the training phase.Our code will be available at:https://gitee.com/yongzhenke/efe-sdg
- 28 October 2025
- Published by Springer Nature in Research Square
Open AccessEFE-SDG: Efficient Feature Extraction of Finetuning-Free Model in Subject-Driven Generation
Text-to-image models mark a major advance in AI, and subject-driven generation techniques hold promise. The subject-driven models that are currently avail- able still necessitate a compromise between the fidelity of the subject and the extent of text-to-image generation. Furthermore, they cannot provide a compre- hensive understanding of the reference image based on a limited amount of data. To address these limitations, we propose the EFE-SDG model, which is divided into three blocks: (i) the MOCI block employs the Gpt-4o and other large mod- els for the processing of the reference image to obtain more information about the reference image. (ii) The ASCA module, based on the decoupling strategy of the cross-attention module, performs additional processing on the reference image to obtain more detailed and richer high-level features of the subject. (iii) The subject Feature Adaptive Attention Rules is used to fuse the low-level fea- tures extracted by the ref-diffusion model and the high-level features extracted by ASCA at the Attention layer of the main-diffusion model. In addition, we use the ref-diffusion model to extract low-level feature inputs from the reference features to the main-diffusion model, which circumvents the different training distributions of the other encoders and the main-diffusion model. In our compar- ative experiments, with respect to graphic alignment, our approach demonstrates performance similar to that of the other methods, but our approach uses sub- stantially smaller datasets and fewer computational resources during the training phase.Our code will be available at:https://gitee.com/yongzhenke/efe-sdg
- 15 October 2025
- Published by Springer Nature in Applied Intelligence
- Vol. 55 (15) , 1-16
DDSPNet: A Two-stage defect detection and severity prediction model for industrial products
- 15 October 2025
- Published by Springer Nature in Applied Intelligence
- Vol. 55 (15) , 1-16
DDSPNet: A Two-stage defect detection and severity prediction model for industrial products
- 21 August 2025
- Published by Springer Nature in Multimedia Systems
- Vol. 31 (5) , 1-11
Aesblip2: generating image aesthetic caption via prompting
- 21 August 2025
- Published by Springer Nature in Multimedia Systems
- Vol. 31 (5) , 1-11
Aesblip2: generating image aesthetic caption via prompting