CV

Research experience

• University of Utah Salt Lake City, UT
  • Research Assistant, Computer Vision Group and Data Group 2016.8 - Present
    • On SLAM
      • Proposed a system to register sparse 3D scans using alternating projection~\cite{ranade2020mapping}
      • Built an image-based localization system for vehicles by aggregating image \ semantic edges~\cite{yu2018vlase}
    • On Network Pruning
      • Studying network pruning with Bayesian variable selection and combinatorial \ optimization methods
      • Built a linear programming system to scale up exact neural network compression \by ReLU stability~\cite{serra2021scaling}
      • Proposed a layer-wise pruning algorithm to achieve strict identity mappings \ in deep residual networks~\cite{yu2018learning}
    • On 3D Reconstruction
      • Proposed a GCN framework for 3D face reconstruction
      • from a single image \ with SOTA performance
    • on Sentiment Analysis
      • Built a framework of spatio-temporal sentiment analysis on tweets to predict \ the US election 2016~\cite{paul2017compass}

• Mitsubishi Electric Research Laboratories Cambridge, MA

  • Research Intern, Computer Vision Group 2019.5 - 2019.11
    • Intern project: 3D Human Shape Recovery from A Single Image with Bilayer-Graph
    • Proposed a deep human shape reconstruction framework with SOTA performance

Education

• Master in Computer Science, University of Utah, 2016 - 2018
• Ph.D in Computer Science, University of Utah, 2018 - Now

Publications

Paul, D., Li, F., Teja, M.K., Yu, X. and Frost, R., 2017, August. " Compass: Spatio temporal sentiment analysis of US election what twitter says!." In Proceedings of the 23rd ACM SIGKDD international conference on knowledge discovery and data mining (pp. 1585-1594). (KDD 2017) Paper

Yu, X., Yu, Z. and Ramalingam, S., 2018. Learning strict identity mappings in deep residual networks. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (pp. 4432-4440) (CVPR 2018) Paper

Yu, X., Chaturvedi, S., Feng, C., Taguchi, Y., Lee, T.Y., Fernandes, C. and Ramalingam, S., 2018, October. Vlase: Vehicle localization by aggregating semantic edges. In 2018 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) (pp. 3196-3203). IEEE. (IROS 2018) Paper

Ranade, S.*, ,Yu, X.*, Kakkar, S., Miraldo, P. and Ramalingam, S., 2020. Mapping of Sparse 3D Data using Alternating Projection. In Proceedings of the Asian Conference on Computer Vision.(ACCV 2020 Oral) Paper

Serra, T., Yu, X., Kumar, A. and Ramalingam, S., 2021. Scaling Up Exact Neural Network Compression by ReLU Stability. arXiv preprint arXiv:2102.07804. (NeurIPS 2021) Paper

Yu X., & Baar J, Chen S. Joint 3D Human Shape Recovery and Pose Estimation from A Single Image with Bilayer-Graph. in International Conference on 3D Vision (3DV 2021) Paper

Yu X.*, Serra T.*, Ramalingam S., Zhe S. The Combinatorial Brain Surgeon: Pruning Weights That Cancel One Another in Neural Networks. In International Conference on Machine Learning (ICML 2022). Paper

Good A.*, Lin J.*, Yu X.*, Sieg H., Ferguson M., Zhe S., Wieczore J., & Serra T. Recall Ditortion in Neural Network Pruning and the Undecayed Pruning Algorithm. Advances in Neural Information Processing Systems (NeurIPS 2022) Paper

Li, S.*, Phillips, J.*, Yu, X., Kirby, R., & Zhe, S. Batch Multi-Fidelity Active Learning with Budget Constraints. Advances in Neural Information Processing Systems (NeurIPS 2022). Paper

Cai J., Nguyen K., Shrestha N., Good A., Tu R., Yu X., & Serra T. Getting away with more network pruning: From sparsity to geometry and linear regions. International Confer- ence on Integration of Constraint Programming, Artificial Intelligence, and Operations Research (CPAIOR 2023) Paper

Fang, S., Yu, X., Li, S., Wang, Z., Kirby R., & Zhe, S. Streaming Factor Trajectory Learning for Temporal Tensor Decomposition. Advances in Neural Information Processing Systems (NeurIPS 2023) Paper

Teaching

Service and leadership

• PC Member / Reviewers for RIOS