Please also see my Google Scholar Profile for the most updated publications.

* indicates equal contribution

Selected Publications

Zhihui Zhu (equal), Tianyu Ding (equal), Jinxin Zhou, Xiao Li, Chong You, Jeremias Sulam, Qing Qu (2021). A Geometric Analysis of Neural Collapse with Unconstrained Features. (NeurIPS’21, spotlight, top 3%)
Preprint – PDF –  Slides –  BibTex – Code – Video

Qing Qu, Yuexiang Zhai, Xiao Li, Yuqian Zhang, Zhihui Zhu (2020). Analysis of the Optimization Landscapes for Overcomplete Representation Learning. ICLR’20, oral, top 1.9%.
Preprint – PDF – Slides – BibTex

Qing Qu, Xiao Li, Zhihui Zhu (2019). Exact Recovery of Multichannel Sparse Blind Deconvolution via Gradient Descent. SIAM Journal on Imaging Science, 13(3): 1630–1652, 2020. (NeurIPS’19, spotlight, top 3%).
Preprint – PDF – Code – Poster – Slides – BibTex

Yuqian Zhang, Qing Qu, John Wright (2020). From Symmetry to Geometry: Tractable Nonconvex Problems. Under Review at Proceedings of IEEE.
Preprint – PDF –  Slides – BibTex

Ju Sun, Qing Qu, John Wright (2018). A Geometric Analysis of Phase Retrieval. Foundations of Computational Mathematics, 18(5):1131–1198, 2018. (ISIT’16).
Preprint – PDF – Code – Slides – BibTex

2023

Huijie Zhang*, Yifu Lu*, Ismail Alkhouri, Saiprasad Ravishankar, Dogyoon Song, Qing Qu (2023). Improving Efficiency of Diffusion Models via Multi-Stage Framework and Tailored Multi-Decoder Architectures. ArXiv Preprint arXiv:2312.09181, 2023.
Preprint – PDF –  BibTex

Soo Min Kwon*, Zekai Zhang*, Dogyoon Song, Laura Balzano, Qing Qu (2023). Efficient Compression of Overparameterized Deep Models through Low-Dimensional Learning Dynamics. ArXiv Preprint arXiv:2311.05061, 2023.
Preprint – PDF –  BibTex – Code

Peng Wang*, Xiao Li*, Can Yaras, Zhihui Zhu, Laura Balzano, Wei Hu, Qing Qu (2023). Understanding Deep Representation Learning via Layerwise Feature Compression and Discrimination. ArXiv Preprint arXiv:2311.02960, 2023.
Preprint – PDF –  BibTex – Code

Pengyu Li*, Yutong Wang*, Xiao Li, Qing Qu (2023). Neural Collapse in Multi-label Learning with Pick-all-label Loss. ArXiv Preprint arXiv:2310.15903, 2023.
Preprint – PDF –  BibTex – Code

Jiachen Jiang*, Jinxin Zhou*, Peng Wang, Qing Qu, Dustin Mixon, Chong You*, Zhihui Zhu* (2023). Generalized Neural Collapse for a Large Number of Classes. ArXiv Preprint arXiv:2310.05351, 2023.
Preprint – PDF –  BibTex

Huijie Zhang*, Jinfan Zhou*, Yifu Lu, Minzhe Guo, Liyue Shen, Qing Qu (2023). The Emergence of Reproducibility and Consistency in Diffusion Models. ArXiv Preprint arXiv:2310.05264, 2023.
Preprint – PDF –  BibTex – Slides

Yuexiang Zhai, Shengbang Tong, Xiao Li, Mu Cai, Qing Qu, Yong Jae Lee, Yi Ma (2023). Investigating the Catastrophic Forgetting in Multimodal Large Language Models. ArXiv Preprint arXiv:2309.10313, 2023.
Preprint – PDF –  BibTex

Ismail Alkhouri*, Shijun Liang*, Rongrong Wang, Qing Qu, Saiprasad Ravishankar (2023). Diffusion-based Adversarial Purification for Robust Deep MRI Reconstruction. ArXiv Preprint arXiv:2309.05794, 2023.
Preprint – PDF –  BibTex

Can Yaras*, Peng Wang*, Wei Hu, Zhihui Zhu, Laura Balzano, Qing Qu (2023). The Law of Parsimony in Gradient Descent for Learning Deep Linear Networks. ArXiv Preprint arXiv:2306.01154, 2023.
Preprint – PDF –  BibTex – Code – Slides

Bowen Song*, Soo Min Kwon*, Zecheng Zhang, Xinyu Hu, Qing Qu, Liyue Shen (2023). Solving Inverse Problems with Latent Diffusion Models via Hard Data Consistency. ArXiv Preprint arXiv:2307.08123, 2023.
Preprint – PDF –  BibTex

Evan Bell, Shijun Liang, Qing Qu, Saiprasad Ravishankar (2023). Robust Self-Guided Deep Image Prior. IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP’23), 2023.
Preprint – PDF –  BibTex

2022

Xiao Li*, Sheng Liu*, Jinxin Zhou, Xinyu Lu, Carlos Fernandez-Granda, Zhihui Zhu, Qing Qu (2023). Principled and Efficient Transfer Learning of Deep Models via Neural Collapse. ArXiv Preprint arXiv:2212.12206, 2022.
Preprint – PDF –  BibTex

Can Yaras*, Peng Wang*, Zhihui Zhu, Laura Balzano, Qing Qu (2022). Neural Collapse with Normalized Features: A Geometric Analysis over the Riemannian Manifold. Neural Information Processing Systems (NeurIPS’22), 2022.
Preprint – PDF –  BibTex – Code

Jinxin Zhou, Chong You, Xiao Li, Kangning Liu, Sheng Liu, Qing Qu, Zhihui Zhu (2022). Are All Losses Created Equal: A Neural Collapse Perspective. Neural Information Processing Systems (NeurIPS’22), 2022.
Preprint – PDF –  BibTex – Code

Tuba Sarwar, Can Yaras, Xiang Li, Qing Qu, Pei-Cheng Ku (2022). Miniaturizing a Chip-Scale Spectrometer Using Local Strain Engineering and Total-Variation Regularized Reconstruction. Nano Letters 22 (20), 8174-8180, 2022.
Preprint – PDF –  BibTex

Shuo Xie, Jiahao Qiu, Ankita Pasad, Li Du, Qing Qu, Hongyuan Mei (2022). Hidden State Variability of Pretrained Language Models Can Guide Computation Reduction for Transfer Learning. Findings of Empirical Methods in Natural Language Processing (EMNLP), 2022.
Preprint – PDF –  BibTex – Code

Sheng Liu, Zhihui Zhu, Qing Qu, Chong You (2022). Robust Training under Label Noise by Over-parameterization. International Conference on Machine Learning (ICML’22), 2022.
Preprint – PDF –  BibTex – Code

Jinxin Zhou*, Xiao Li*, Tianyu Ding, Chong You, Qing Qu*, Zhihui Zhu* (2022). On the Optimization Landscape of Neural Collapse under MSE Loss: Global Optimality with Unconstrained Features. International Conference on Machine Learning (ICML’22), 2022.
Preprint – PDF –  BibTex – Code

Peng Wang*, Huikang Liu*, Can Yaras*, Laura Balzano, Qing Qu (2022). Linear Convergence Analysis of Neural Collapse with Unconstrained Features. OPT 2022: Optimization for Machine Learning (NeurIPS 2022 Workshop), 2022.
Preprint – PDF –  BibTex

2021

Lijun Ding*, Liwei Jiang*, Yudong Chen, Qing Qu, Zhihui Zhu (2021). Rank Overspecified Robust Matrix Recovery: Subgradient Method and Exact Recovery. Neural Information Processing Systems (NeurIPS’21), 2021.
Preprint – PDF –  BibTex

Zhihui Zhu*, Tianyu Ding*, Jinxin Zhou, Xiao Li, Chong You, Jeremias Sulam, Qing Qu (2021). A Geometric Analysis of Neural Collapse with Unconstrained Features. Neural Information Processing Systems (NeurIPS’21), 2021. (spotlight, top 3%)
Preprint – PDF –  Slides –  BibTex – Code

Sheng Liu*, Xiao Li*, Yuexiang Zhai, Chong You, Zhihui Zhu, Carlos Fernandez-Granda, Qing Qu (2021). Convolutional Normalization: Improving Deep Convolutional Network Robustness and Training. Neural Information Processing Systems (NeurIPS’21), 2021.
Preprint – PDF –  BibTex – Code

Xiao Li*, Shixiang Chen*, Zengde Deng, Qing Qu, Zhihui Zhu, Anthony Man Cho So (2021). Weakly Convex Optimization over Stiefel Manifold Using Riemannian Subgradient-Type Methods. SIAM Journal on Optimization, 31(3): 1605-1634, 2021.
Preprint – PDF – Code – BibTex

2020

Qing Qu, Yuexiang Zhai, Xiao Li, Yuqian Zhang, and Zhihui Zhu (2020). Geometric Analysis of Nonconvex Optimization Landscapes for Overcomplete Learning. International Conference on Learning Representations (ICLR’20), 2020. (oral, top 1.9%)
Preprint – PDF – Slides – BibTex

Chong You*, Zhihui Zhu*, Qing Qu, Yi Ma (2020). Robust Recovery via Implicit Bias of Discrepant Learning Rates for Double Over-parameterization. Neural Information Processing Systems (NeurIPS’20), 2020. (spotlight, top 4%)
Preprint – PDF – BibTex – Code

Yenson Lau*, Qing Qu*, Han-wen Kuo, Pengcheng Zhou, Yuqian Zhang, and John Wright (2020). Short and Sparse Deconvolution — A Geometric Approach. International Conference on Learning Representations (ICLR’20), 2020.
Preprint – PDF – Code – Poster – Slides – Website – BibTex

Yuqian Zhang, Qing Qu, John Wright (2020). From Symmetry to Geometry: Tractable Nonconvex Problems. In Submission to Proceedings of IEEE.
Preprint – PDF –  Slides – BibTex

Qing Qu*, Zhihui Zhu*, Xiao Li, Manolis C. Tsakiris, John Wright, Rene Vidal (2020). Finding the Sparsest Vectors in a Subspace: Theory, Algorithms, and Applications. In Submission.
Preprint – PDF – Slides – BibTex

2019 & Prior

Qing Qu, Xiao Li, Zhihui Zhu (2019). Exact Recovery of Multichannel Sparse Blind Deconvolution via Gradient Descent. SIAM Journal on Imaging Science, 13(3): 1630–1652, 2020. (NeurIPS’19, spotlight, top 3%).
Preprint – PDF – Code – Poster – Slides – BibTex

Qing Qu, Yuqian Zhang, Yonina C. Eldar, John Wright (2019). Convolutional Phase Retrieval via Gradient Descent. IEEE Trans. on Information Theory, 66(3):1785–1821, Mar. 2020. (NeurIPS’17).
Preprint – PDF – Code – Poster – Slides – BibTex

Ju Sun, Qing Qu, John Wright (2018). A Geometric Analysis of Phase Retrieval. Foundations of Computational Mathematics, 18(5):1131–1198, 2018. (ISIT’16).
Preprint – PDF – Code – Slides – BibTex

Ju Sun, Qing Qu, John Wright (2017). Complete Dictionary Recovery over the Sphere I: Overview and the Geometric Picture. IEEE Trans. on Information Theory, 63(2): 853 – 884, Feb. 2017. (ICML’15, best student paper award for SPARS’15).
Preprint – PDF – Code – Poster – BibTex

Ju Sun, Qing Qu, John Wright (2017). Complete Dictionary Recovery over the Sphere II: Recovery by Riemannian Trust-region Method. IEEE Trans. on Information Theory, 63(2): 853 – 884, Feb. 2017.
Preprint – PDF – Code – Poster – BibTex

Qing Qu, Ju Sun, John Wright (2016). Finding a Sparse Vector in a Subspace: Linear Sparsity Using Alternating Directions. IEEE Trans. on Information Theory, 62(10): 5855 – 5880, Oct. 2016 (NeurIPS’14).
Preprint – PDF – Code – Poster – Slides – BibTex

Qing Qu, Nasser M. Nasrabadi, Trac D. Tran (2015). Subspace Vertex Pursuit: A Fast and Robust Near-Separable Nonnegative Matrix Factorization Method for Hyperspectral Unmixing. IEEE Journal of Selected Topics in Signal Processing, 9(6): 1142 – 1155, Sept. 2015. (ICASSP’14)
PDF – Poster – BibTex

Ju Sun, Qing Qu, John Wright (2015). When Are Nonconvex Problems Not Scary?. NeurIPS’15 Workshop on Nonconvex Optimization for Machine Learning, 2015.
Preprint – PDF –  BibTex

Qing Qu, Nasser M. Nasrabadi, Trac D. Tran (2014). Abundance Estimation for Bilinear Mixture Models via Joint Sparse and Low-Rank Representation. IEEE Transactions on Geoscience and Remote Sensing, 52(7): 4404 – 4423, Jul. 2014. (ICASSP’13)
PDF – Code – Poster – Slides – BibTex

Xiaoxia Sun, Qing Qu, Nasser M. Nasrabadi, Trac D. Tran (2014). Structured Priors for Sparse-Representation-Based Hyperspectral Image Classification. IEEE Geoscience and Remote Sensing Letters, 11(7): 1235 – 1239, Jul. 2014.
Preprint – PDF – BibTex

Jian Jin, Qing Qu, Yuantao Gu (2013). Robust Zero-point Attraction Least Mean Square Algorithm on Near Sparse System Identification. IET Signal Processing, 7(3): 210 – 218, May 2013.
Preprint – PDF – BibTex