Neural étendue expander for ultra-wide-angle high-fidelity holographic display

Area of Science:

• Optics and Photonics
• Computer Vision
• Display Technology

Background:

• Holographic displays use spatial light modulators to create light fields for virtual and augmented reality.
• Limited spatial resolution restricts diffraction angles, resulting in low étendue (display area × solid angle).
• Low étendue necessitates trade-offs between field-of-view (FOV) and display size in current holographic systems.

Purpose of the Study:

• To overcome the étendue limitation in holographic displays.
• To enable ultra-wide FOV holographic displays without sacrificing display size or image fidelity.
• To introduce a new class of optical elements: neural étendue expanders.

Main Methods:

• Developed neural étendue expanders, a novel optical element type.
• Trained these elements using a natural image dataset.
• Experimentally integrated neural étendue expanders into a holographic display system.

Main Results:

• Achieved a 64× étendue expansion for full-color natural images.
• Expanded the horizontal and vertical FOV by an order of magnitude.
• Maintained high-fidelity reconstruction quality, exceeding 29 dB PSNR for retinal-resolution images.

Conclusions:

• Neural étendue expanders effectively lift the étendue limitation in holographic displays.
• This technology enables significantly wider FOVs and maintains high image quality.
• Paves the way for more immersive and capable virtual and augmented reality experiences.