A Lightweight Complex-Valued Deformable CNN for High-Quality Computer-Generated Holography

TL;DR

This paper introduces a lightweight complex-valued deformable CNN that enhances feature extraction for high-quality computer-generated holography, achieving state-of-the-art results with fewer parameters.

Contribution

It proposes a novel complex-valued deformable convolution that dynamically adjusts the receptive field, improving hologram reconstruction accuracy with a more efficient model.

Findings

Achieves higher PSNR than existing models in simulated and optical tests.

Reduces model parameters to about one-eighth of comparable methods.

Surpasses state-of-the-art performance in holographic reconstruction quality.

Abstract

Holographic displays have significant potential in virtual reality and augmented reality owing to their ability to provide all the depth cues. Deep learning-based methods play an important role in computer-generated holography (CGH). During the diffraction process, each pixel exerts an influence on the reconstructed image. However, previous works face challenges in capturing sufficient information to accurately model this process, primarily due to the inadequacy of their effective receptive field (ERF). Here, we designed complex-valued deformable convolution for integration into network, enabling dynamic adjustment of the convolution kernel's shape to increase flexibility of ERF for better feature extraction. This approach allows us to utilize a single model while achieving state-of-the-art performance in both simulated and optical experiment reconstructions, surpassing existing open-source models. Specifically, our method has a peak signal-to-noise ratio that is 2.04 dB, 5.31 dB, and 9.71 dB higher than that of CCNN-CGH, HoloNet, and Holo-encoder, respectively, when the resolution is 1920$\times$1072. The number of parameters of our model is only about one-eighth of that of CCNN-CGH.