Accurate and timely channel state information (CSI) is essential for the performance of next-generation wireless systems, particularly those operating in high-frequency bands with large-scale antenna arrays. These systems, including future 6G networks, rely on spatially resolved CSI to support tasks such as beamforming, mobility management, and interference mitigation. However, acquiring reliable CSI in practical environments remains challenging due to high measurement cost, environmental complexity, and real-time constraints. This project develops a modeling framework that integrates limited radio measurements with spatial priors derived from environmental sensing. Specifically, the project investigates how geometric and visual information can be used to infer signal behavior in environments with constrained sensing capability. Rather than introducing a new channel abstraction, the project focuses on applying radiance-inspired modeling to characterize local electromagnetic behavior as a function of position and direction. The resulting models aim to support compact, data-efficient CSI reconstruction for structured scenarios such as indoor or urban deployments. Broader impacts include the integration of project outcomes into advanced wireless curriculum and engagement of students through interdisciplinary research. Project data, code, and validation tools will be open-sourced to support reproducibility and research dissemination. The proposed research explores a data fusio