Faculty Profile

Xi Peng

Xi Peng

Assistant Professor
440 Smith Hall
Newark, Delaware 19716
P: 302-831-3185
xipeng@udel.edu

Personal Website



EDUCATION

PhD | 2018 | Rutgers, The State University of New Jersey
MS | 2011 | Chinese Academy of Sciences, Beijing, China
BS | 2008 | Beihang University, Beijing, China

SHORT BIO

Dr. Xi Peng works in the area of Deep Learning, Machine Learning, and Computer Vision. His research interests lie in Structural and Model-Oriented Deep Learning. The goal is to achieve Robust & Explainable & Efficient AI systems for intelligent data analytics:

1) Can we learn AI systems promptly and adequately when only limited or skewed data are available? (Efficient AI)
2) Can AI system make robust and safe decisions when training and testing data follow different distributions? (Robust AI)
3) Can we extend from curve-fitting to model causal inference between data and environment? (Explainable AI)

More: Google Scholar   Deep-REAL lab

SELECT PUBLICATIONS

[1] Z. Tang, X. Peng*, K. Li, D. Metaxas. “Towards Efficient U-Nets: A Coupled and Quantized Approach.” IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI, impact factor: 17.73), pages 1–13, 2019. (*corresponding author)

[2] R. Mehrizi, X. Peng, X. Xu, S. Zhang, K. Li. Predicting 3D Lower-Back Joint Load in Lifting: A Deep Pose Estimation Approach.” IEEE Transactions on Human-Machine System (THMS), 2019.

[3] Y. Tian*, L. Zhao*, X. Peng, D. Metaxas. “Rethinking Kernel Methods for Node Representation Learning on Graphs.” In Proceedings of Conference on Neural Information Processing Systems (NeurIPS), 2019. (*contributed equally)

[4] Y. Zhu, J. Xie, Z. Tang, X. Peng, A. Elgammal. Semantic-Guided Multi-Attention Localization for Zero-Shot Learning.” In Proceedings of Conference on Neural Information Processing Systems (NeurIPS), 2019.

[5] Z. Tang, X. Peng, T. Li, Y. Zhu, D. Metaxas. “AdaTransform: Adaptive Data Transformation.” In Proceedings of the IEEE International Conference on Computer Vision (ICCV, oral, acceptance rate: 4.7%), 2019.

[6] L. Zhao, X. Peng, Y. Tian, M. Kapadia, D. Metaxas. “Semantic Graph Convolutional Networks for 3D Human Pose Regression.” In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2019.

[7] X. Peng, Z. Tang, F. Yang, R. Feris, D. Metaxas. “Jointly Optimize Data and Network Training: Adversarial Data Augmentation.” In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2018.

[8] X. Peng, X. Yu, K. Sohn, D. Metaxas, M. Chandraker. “Reconstruction-based Disentanglement for Pose-invariant Face Recognition.” In Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2017.

[9] X. Peng, R. Feris, X. Wang, D. Metaxas. “A Recurrent Encoder-Decoder Network for Sequential Face Alignment.” In European Conference on Computer Vision (ECCV, Oral, best student paper runner-up, 6/1500+), 2016.

[10] X. Peng, S. Zhang, Y. Yang, D. Metaxas. PIEFA: Personalized Incremental and Ensemble Face Alignment.” In Proceedings of the IEEE International Conference on Computer Vision (ICCV), 2015.

 View Full CV

People Faculty Profile