From Less to More: Spectral Splitting and Aggregation Network for Hyperspectral Face Super-Resolution

TL;DR

This paper introduces SSANet, a spectral splitting and aggregation network designed for hyperspectral face super-resolution, effectively handling limited training data by exploiting spectral correlations and data augmentation.

Contribution

The paper proposes a novel spectral splitting and aggregation strategy (SSAS) and a self-representation based data expansion method for hyperspectral face super-resolution with limited samples.

Findings

SSANet effectively models spatial and spectral correlations.

Spectral splitting improves training efficiency with limited data.

Data augmentation alleviates small sample size issues.

Abstract

High-resolution (HR) hyperspectral face image plays an important role in face related computer vision tasks under uncontrolled conditions, such as low-light environment and spoofing attacks. However, the dense spectral bands of hyperspectral face images come at the cost of limited amount of photons reached a narrow spectral window on average, which greatly reduces the spatial resolution of hyperspectral face images. In this paper, we investigate how to adapt the deep learning techniques to hyperspectral face image super-resolution (HFSR), especially when the training samples are very limited. Benefiting from the amount of spectral bands, in which each band can be seen as an image, we present a spectral splitting and aggregation network (SSANet) for HFSR with limited training samples. In the shallow layers, we split the hyperspectral image into different spectral groups. Then, we gradually aggregate the neighbor bands at deeper layers to exploit spectral correlations. By this spectral splitting and aggregation strategy (SSAS), we can divide the original hyperspectral image into multiple samples (\emph{from less to more}) to support the efficient training of the network and effectively exploit the spectral correlations among spectrum. To cope with the challenge of small training sample size (S3) problem, we propose to expand the training samples by a self-representation model and symmetry-induced augmentation. Experiments show that SSANet can well model the joint correlations of spatial and spectral information. By expanding the training samples, SSANet can effectively alleviate the S3 problem.