Holography has immense potential for near-eye displays in virtual and augmented reality (VR/AR), providing natural 3D depth cues through wavefront reconstruction. However, balancing the field of view (FOV) with the eyebox remains challenging, constrained by the étendue limitation. Additionally, holographic image quality is often compromised due to differences between actual wave propagation and simulation models. This study addresses these by expanding the eyebox via multi-angle illumination, and enhancing image quality with end-to-end pupil-aware hologram optimization. Further, energy efficiency is improved by incorporating higher-order diffractions and pupil constraints. We explore a Pupil-HOGD algorithm for multi-angle illumination and validate it with a dual-angle holographic display prototype. Integrated with camera calibration and tracked eye position, the developed Pupil-HOGD algorithm improves image quality and expands the eyebox by 50% horizontally. We envision this approach extends the space-bandwidth product (SBP) of holographic displays, enabling broader applications in immersive, high-quality visual computing.

With the development of adversarial attack in deep learning, it is critical for abnormal detector to not only discover the out-of-distribution samples but also provide defence against the adversarial attacker. Since few previous universal detector is known to work well on both tasks, we consider against both scenarios by constructing a robust and effective technique, where one sample could be regarded as the abnormal sample if it exhibits a higher image reconstruction error. Due to the training instability issues existed in previous generative adversarial networks (GANs) based methods, in this paper we propose a dual auxiliary autoencoder to make a tradeoff between the capability of generator and discriminator, leading to a more stable training process and high-quality image reconstruction. Moreover, to generate discriminative and robust latent representations, the mutual information estimator regarded as latent regularizer is adopted to extract the most unique information of target class. Overall, our generative dual adversarial network simultaneously optimizes the image reconstruction space and latent space to improve the performance. Experiments show that our model has the clear superiority over cutting edge semi-supervised abnormal detectors and achieves the state-of-the-art results on the datasets.