Compressive Imaging via Approximate Message Passing скачать в хорошем качестве

Compressive Imaging via Approximate Message Passing

This tutorial video describes some of our research work on compressive imaging, which relies heavily on the approximate message passing (AMP) framework for solving linear inverse problems. Two algorithms are discussed. The first reconstructs images from noisy linear measurements; applications could be in medical imaging or radio astronomy, to name a few. The second algorithm reconstructs image cubes acquired by compressive hyper spectral hardware systems, for example the well-known CASSI system. Both algorithms reconstruct quite well, are robust to measurement noise, and are faster than other competing algorithms. More details about this work appear in our papers, which are listed at the end of the video.