Petros T. Boufounos
Mitsubishi Electric Research Laboratories
250 Papers
2K Citations
Petros T. Boufounos is an academic researcher from Mitsubishi Electric Research Laboratories. The author has contributed to research in topics: Compressed sensing & Signal. The author has an hindex of 31, co-authored 223 publications. Previous affiliations of Petros T. Boufounos include Mitsubishi Electric & Chalmers University of Technology.
Chat about Author
Papers
Dictionary learning based pan-sharpening
Dehong Liu,Petros T. Boufounos +1 more
- 25 Mar 2012
TL;DR: This paper proposes a dictionary learning based pan-sharpening process to reduce the color distortion caused by the interpolation of the MS imagery, and generates an improved MS image which is sparse with respect to a dictionary learned from the image data.
Radar Autofocus Using Sparse Blind Deconvolution
Hassan Mansour,Dehong Liu,Petros T. Boufounos,Ulugbek S. Kamilov +3 more
- 15 Apr 2018
TL;DR: An alternating minimization framework that leverages the sparsity and piece-wise smoothness of the radar scene, as well as the one-sparse property of the two dimensional shift kernels for each antenna measurement is developed.
17
Extended Object Tracking With Automotive Radar Using B-Spline Chained Ellipses Model
Gang Yao,Pu Wang,Karl Berntorp,Hassan Mansour,Petros T. Boufounos,Philip Orlik +5 more
- 06 Jun 2021
TL;DR: In this paper, a B-spline chained ellipses model representation for extended object tracking (EOT) using high-resolution automotive radar measurements is introduced, and the proposed model parameters are learned using the expectation-maximization (EM) algorithm.
17
Sparsity-driven distributed array imaging
Dehong Liu,Ulugbek S. Kamilov,Petros T. Boufounos +2 more
- 01 Dec 2015
TL;DR: This work develops compressive sensing based methods to improve imaging performance, and imposes sparsity on the complex-valued reconstruction of the region of interest, with the non-zero coefficients corresponding to the imaged targets.
17
Quantized embeddings: an efficient and universal nearest neighbor method for cloud-based image retrieval
TL;DR: A rate-efficient, feature-agnostic approach for encoding image features for cloud-based nearest neighbor search, using quantized random projections of the image features under consideration, and showing that pair-wise distances between the underlying feature vectors are preserved in the corresponding quantized embeddings.