Xin Yu

I am a final-year Ph.D. student in Kahlert School of Computing at the University of Utah. I’m now doing research on algorithm and data efficiency , under the supervision of Prof. Shandian Zhe and Prof. Srikumar Ramalingam.

My research interests lie in the intersection of data/algorithm efficiency and Optimization. Specifically, I’m exploring efficient computation, economic data abstraction and collection, and bias reduction and robustness enhancement in algorithms. Drawing upon my expertise in combinatorial optimization, and statistical analysis, I aim to address these multifaceted challenges of algorithm deployment in resource-limited devices.

News

[01/2024]: Our paper Multi-Resolution Active Learning of Fourier Neural Operators is accepted by AISTATS 2024 (oral).
[01/2024]: Our paper Functional Bayesian Tucker Decomposition for Continuous-indexed Tensor Data is accepted by ICLR 2024.
[12/2023]: Our paper Streaming Factor Trajectory Learning for Temporal Tensor Decomposition will be presented on NeurIPS 2023.
[05/2023]: Our paper Getting away with more network pruning: From sparsity to geometry and linear regions will be presented on CPAIOR 2023.
[11/2022]: Our papers Batch Multi-Fidelity Active Learning with Budget Constraints and Recall Distortion in Neural Network Pruning and the Undecayed Pruning Algorithm are presented on NeurIPS 2022.
[07/2022]: Our paper The Combinatorial Brain Surgeon: Pruning Weights That Cancel One Another in Neural Networks wil be presented on ICML 2022.
[12/2021]: Our paper Joint 3D Human Shape Recovery and Pose Estimation from a Single Image with Bilayer Graph wil be presented on 3DV 2021.
[12/2021]: Our paper Scaling Up Exact Neural Network Compression by ReLU Stability wil be presented to NeurIPS 2021.
[11/2020]: Our paper Mapping of Sparse 3D Data using Alternating Projection wil be presented on ACCV 2020 (oral).
[10/2018]: Our paper VLASE: Vehicle localization by aggregating semantic edges wil be presented on IROS 2018.
[06/2018]: Our paper Learning strict identity mappings in deep residual networks wil be presented on CVPR 2018.

Education

Publications

Teaching

Service and leadership