Limited availability of labeled multi-spectral datasets and high-dynamic range information restricts accurate facial attribute classification. Overcoming data scarcity enables robust recognition across diverse lighting and spectral conditions.