Complex spatial light modulation capability of a dual layer in-plane switching liquid crystal panel

Waveguide holography for 3D augmented reality glasses

Near-eye displays are fundamental technology in the next generation computing platforms for augmented reality and virtual reality. However, there are remaining challenges to deliver immersive and comfortable visual experiences to users, such as compact form factor, solving vergence-accommodation conflict, and achieving a high resolution with a large eyebox. Here we show a compact holographic near-eye display concept that combines the advantages of waveguide displays and holographic displays to overcome the challenges towards true 3D holographic augmented reality glasses. By modeling the coherent light interactions and propagation via the waveguide combiner, we demonstrate controlling the output wavefront using a spatial light modulator located at the input coupler side. The proposed method enables 3D holographic displays via exit-pupil expanding waveguide combiners, providing a large software-steerable eyebox. It also offers additional advantages such as resolution enhancement capability by suppressing phase discontinuities caused by pupil replication process. We build prototypes to verify the concept with experimental results and conclude the paper with discussion.

Real-time 4K computer-generated hologram based on encoding conventional neural network with learned layered phase

Learning-based computer-generated hologram (CGH) demonstrates great potential for real-time high-quality holographic displays. However, real-time 4K CGH generation for 3D scenes remains a challenge due to the computational burden. Here, a variant conventional neural network (CNN) is presented for CGH encoding with learned layered initial phases for layered CGH generation. Specifically, the CNN predicts the CGH based on the input complex amplitude on the CGH plane, and the learned initial phases act as a universal phase for any target images at the target depth layer. These phases are generated during the training process of the coding CNN to further optimize the quality. The CNN is trained to learn encoding 3D CGH by randomly selecting the depth layer in the training process, and contains only 938 parameters. The generation time for a 2D 4K CGH is 18 ms, and is increased by 12 ms for each layer in a layered 3D scene. The average Peak Signal to Noise Ratio (PSNR) of each layer is above 30dB in the depth range from 160 to 210 mm. Experiments verify that our method can achieve real-time layered 4K CGH generation.