主要研究方向：

主要研究方向涉及：人工智能，机器学习，数据挖掘，模式识别

具体方向：多模态学习，聚类，鲁棒特征提取，图表征学习，联邦学习

主要研究成果：

（一）基于图卷积的多模态属性社交图表示模型：

1.多视图图卷积

Jiafeng Cheng, Qianqian Wang, Zhiqiang Tao, De-Yan Xie, Quanxue Gao: Multi-View Attribute Graph Convolution Networks for Clustering. IJCAI 2020: 2973-2979

Zihao Zhang, Qianqian Wang, Zhiqiang Tao, Quanxue Gao, Wei Feng，Dropping Pathways Towards Deep Multi-View Graph Subspace Clustering Networks，ACM MM, 3259-3267 (2023)

（二）多模态一致表征：

1.深度多模态聚类

Qianqian Wang,  Jiafeng Cheng, Quanxue Gao,  Guoshuai Zhao, Licheng Jiao: Deep Multi-View Subspace Clustering With Unified and Discriminative Learning. IEEE Trans. Multim. 23: 3483-3493 (2021)

Qianqian Wang, Zhiqiang Tao, Quanxue Gao, Licheng Jiao:Multi-View Subspace Clustering via Structured Multi-Pathway Network. IEEE Trans. Neural Networks Learn. Syst. 35(5): 7244-7250 (2024)

2.缺失多视图聚类

Qianqian Wang, Zhengming Ding, Zhiqiang Tao, Quanxue Gao, Yun Fu: Generative Partial Multi-View Clustering With Adaptive Fusion and Cycle Consistency. IEEE Trans. Image Process. 30: 1771-1783 (2021)

Qianqian Wang, Huanhuan Lian, Gan Sun, Quanxue Gao, Licheng Jiao: iCmSC: Incomplete Cross-Modal Subspace Clustering. IEEE Trans. Image Process. 30: 305-317 (2021)

（三）鲁棒判别表征：

1.基于L2p-范数的鲁棒主成分分析（L2p-PCA）

Qianqian Wang, Quanxue Gao, Xinbo Gao, Feiping Nie: ℓ2, p -Norm Based PCA for Image Recognition. IEEE Trans. Image Process. 27(3): 1336-1346 (2018)

2.鲁棒角度主成分模型(Angle-PCA)

Qianqian Wang, Quanxue Gao, Xinbo Gao, Feiping Nie: Angle Principal Component Analysis. IJCAI 2017: 2936-2942