Christopher Metzler

Assistant Professor of Computer Science

University of Maryland 


Chris is an Assistant Professor in the Department of Computer Science at UMD. He is a member of UMIACS and has a courtesy appointment in the Electrical and Computer Engineering Department. His research develops new systems and algorithms for solving problems in computational imaging, machine learning, and wireless communications. His work has received multiple best paper awards; he recently received an AFOSR Young Investigator Program Award; and he was an Intelligence Community Postdoctoral Research Fellow, an NSF Graduate Research Fellow, a DoD NDSEG Fellow, and a NASA Texas Space Grant Consortium Fellow.

Prospective Students: Students interested in working with me should apply to the UMD CS PhD program and mention my name in their research statement. I'm especially interested in working with students who want to apply machine learning to real-world optical and RF hardware.

Prospective Postdocs: I am not hiring postdocs at this time, but feel free to reach out.


Experience & Education

  • Computational Imaging
  • Machine Learning
  • Statistical Signal Processing


Recent Publications

TurbuGAN: An Adversarial Learning Approach to Spatially-Varying Multiframe Blind Deconvolution with Applications to Imaging Through Turbulence
B. Feng, M. Xie, and C. Metzler. In Preparation 2022.

D-VDAMP: Denoising-Based Approximate Message Passing for Compressive MRI
C. Metzler and G. Wetzstein. ICASSP 2021.

Depth from Defocus with Learned Optics for Imaging and Occlusion-aware Depth Estimation
H. Ikoma, C. Nguyen, C. Metzler, Y. Peng, G. Wetzstein. ICCP 2021.

Keyhole Imaging: Non-Line-of-Sight Imaging and Tracking of Moving Objects Along a Single Optical Path
C. Metzler, D. Lindell, G. Wetzstein. IEEE Transactions on Computational Imaging 2021.

Deep Learning Techniques for Inverse Problems in Imaging
G. Ongie, A. Jalal, C. Metzler, R. Baraniuk, A. Dimakis, R. Willett. IEEE Journal on Selected Areas in Information Theory 2020.

Deep-inverse Correlography: Towards Real-Time High-Resolution Non-Line-of-Sight Imaging
C. Metzler, F. Heide, P. Rangarajan, M. Balaji, A. Viswanath, A. Veeraraghavan, and R. Baraniuk. Optica 2020.

Deep Optics for Single-shot High-dynamic-range Imaging
C. Metzler, H. Ikoma, Y. Peng, G. Wetzstein. CVPR 2020 (Oral).


A Single Laser Fired Through a Keyhole Can Expose Everything Inside a Room
Gizmodo, September 2021.

Seeing Around Corners with Lasers—and Speckle
IEEE Spectrum, January 2020. Also Times of London and The Telegraph.

Software and Datasets

Learned D-AMP, D-AMP, & D-prGAMP Toolbox: Neural networks and algorithms for compressive sensing and compressive phase retrieval. Includes code to train with SURE loss, instead of MSE.

Keyhole Imaging: Code and data demonstrating non-line-of-sight imaging along a single optical path.

D-VDAMP: Variable-density sampled compressive MRI reconstruction code.

Transmission Matrix Dataset: A public dataset for testing phase retrieval algorithms.

Deep Simultaneous Source Separation and Phase Retrieval: Code to solve S3PR using deep generative models.

Deep-Inverse Correlography: Neural networks for imaging around corners using phase retrieval.

Deep Optics for Single-shot HDR: Code and data demonstrating single-shot HDR with learned optics.

Depth from Defocus with Learned Optics: Code and data demonstrating monocular depth estimation with learned optics.

prDeep: A neural-network-based noise-robust phase retreival algorithm.