EE Seminar: Guided Lensless Polarization Imaging
הרישום לסמינר יבוצע בתחילת הסמינר באמצעות סריקת הברקוד למודל (יש להיכנס לפני כן למודל, לא באמצעות האפליקציה)
Registration to the seminar is done at the beginning of the seminar by scanning the barcode for the Moodle (Please enter ahead to the Moodle, NOT by application)
Electrical Engineering Systems Seminar
Speaker: Noa Kraicer
M.Sc. student under the supervision of Prof. Raja Giryes
Monday, 11th May 2026, at 15:00
Room 011, Kitot Building, Faculty of Engineering
Guided Lensless Polarization Imaging
Abstract
Polarization imaging captures information invisible to the human eye and is useful for applications such as biomedical imaging, autonomous driving, and remote sensing. However, conventional polarization cameras are often expensive, bulky, or both, limiting their practical use. In this work, we address compact polarization imaging with a lensless camera, where simple optical coding elements replace the lens and the image is recovered computationally.
We consider a lensless polarization camera composed of a diffuser and a striped polarization mask, which enables the recovery of multi-angle polarization images from a single snapshot. Since lensless polarization reconstruction is highly ill-posed and often suffers from degraded image quality, we improve the reconstruction by using an auxiliary RGB image as structural guidance. Our method follows a two-stage pipeline: physics-based inversion produces an initial reconstruction, followed by a Transformer-based RGB-guided refinement network.
Our results show that the proposed RGB-guided approach substantially improves lensless polarization reconstruction on simulated datasets and generalizes across imaging conditions. We further validate the method on a physical prototype, achieving high-quality real-world reconstructions without fine-tuning.
-סמינר זה ייחשב כסמינר שמיעה לתלמידי תואר שני ושלישי-
This Seminar Is Considered A Hearing Seminar For Msc/Phd Students-
