Unrolled Primal-Dual Networks for Lensless Cameras

TL;DR

This paper introduces a learned primal-dual reconstruction method for lensless cameras that accounts for optical aberrations and depth variations, significantly improving image quality over traditional models.

Contribution

It demonstrates that embedding learnable forward and adjoint models in a primal-dual framework enhances reconstruction quality and introduces a proof-of-concept lensless camera prototype.

Findings

Reconstructed images improved by +5dB PSNR using the proposed method.

The learned model matches state-of-the-art image quality without large network capacity.

Extensive evaluation on open and prototype datasets confirms effectiveness.

Abstract

Conventional image reconstruction models for lensless cameras often assume that each measurement results from convolving a given scene with a single experimentally measured point-spread function. These image reconstruction models fall short in simulating lensless cameras truthfully as these models are not sophisticated enough to account for optical aberrations or scenes with depth variations. Our work shows that learning a supervised primal-dual reconstruction method results in image quality matching state of the art in the literature without demanding a large network capacity. This improvement stems from our primary finding that embedding learnable forward and adjoint models in a learned primal-dual optimization framework can even improve the quality of reconstructed images (+5dB PSNR) compared to works that do not correct for the model error. In addition, we built a proof-of-concept lensless camera prototype that uses a pseudo-random phase mask to demonstrate our point. Finally, we share the extensive evaluation of our learned model based on an open dataset and a dataset from our proof-of-concept lensless camera prototype.