Simulation of Infra-Red (IR), Synthetic Aperture Radar (SAR), and Out- The-Window (OTW) visual imagery plays an important role in the planning and rehearsal of missions and personnel training. The challenge is to develop database and image generation systems that extremely rapidly and in real time process geo-specific Multi- Spectral Imagery (MSI) over large areas into simulated sensor imagery to achieve high real-world accuracy and sensor/OTW-visual correlation. To meet this challenge, a novel architecture called a Neural Network Look-Up Table (NNLUT) which implements spectral conversion by neural networks has been developed. The NNLUT processor is described and examples of highly correlated IR and SAR imagery simulated in real time from MSI by the NNLUT are demonstrated.