Learnable Fractional Reaction-Diffusion Dynamics for Under-Display ToF Imaging and Beyond
This work addresses depth quality issues in under-display ToF imaging for applications like smartphones or augmented reality, representing an incremental advancement by combining neural networks with physical modeling.
The paper tackles the problem of accurate depth sensing in under-display ToF imaging, which suffers from severe degradations like signal attenuation and multi-path interference due to transparent OLED layers, and proposes a hybrid framework called Learnable Fractional Reaction-Diffusion Dynamics (LFRD2) that achieves improved depth restoration, as demonstrated through experiments on four benchmark datasets.
Under-display ToF imaging aims to achieve accurate depth sensing through a ToF camera placed beneath a screen panel. However, transparent OLED (TOLED) layers introduce severe degradations-such as signal attenuation, multi-path interference (MPI), and temporal noise-that significantly compromise depth quality. To alleviate this drawback, we propose Learnable Fractional Reaction-Diffusion Dynamics (LFRD2), a hybrid framework that combines the expressive power of neural networks with the interpretability of physical modeling. Specifically, we implement a time-fractional reaction-diffusion module that enables iterative depth refinement with dynamically generated differential orders, capturing long-term dependencies. In addition, we introduce an efficient continuous convolution operator via coefficient prediction and repeated differentiation to further improve restoration quality. Experiments on four benchmark datasets demonstrate the effectiveness of our approach. The code is publicly available at https://github.com/wudiqx106/LFRD2.