Hyperspectral Benchmark: Bridging the Gap between HSI Applications through Comprehensive Dataset and Pretraining
This addresses a recurring challenge in HSI for researchers and practitioners by providing a standardized benchmark to evaluate and improve methods across diverse applications, though it is incremental as it builds on existing data and pretraining concepts.
The paper tackles the problem of limited dataset sizes in hyperspectral imaging (HSI) applications by introducing a comprehensive benchmark dataset covering food inspection, remote sensing, and recycling, which enables better model assessment and facilitates the development of a pretraining pipeline to improve training stability for larger models.
Hyperspectral Imaging (HSI) serves as a non-destructive spatial spectroscopy technique with a multitude of potential applications. However, a recurring challenge lies in the limited size of the target datasets, impeding exhaustive architecture search. Consequently, when venturing into novel applications, reliance on established methodologies becomes commonplace, in the hope that they exhibit favorable generalization characteristics. Regrettably, this optimism is often unfounded due to the fine-tuned nature of models tailored to specific HSI contexts. To address this predicament, this study introduces an innovative benchmark dataset encompassing three markedly distinct HSI applications: food inspection, remote sensing, and recycling. This comprehensive dataset affords a finer assessment of hyperspectral model capabilities. Moreover, this benchmark facilitates an incisive examination of prevailing state-of-the-art techniques, consequently fostering the evolution of superior methodologies. Furthermore, the enhanced diversity inherent in the benchmark dataset underpins the establishment of a pretraining pipeline for HSI. This pretraining regimen serves to enhance the stability of training processes for larger models. Additionally, a procedural framework is delineated, offering insights into the handling of applications afflicted by limited target dataset sizes.