Computational imaging techniques—exemplified by multi-dimensional light-field encoding and compressive sensing—have broken the traditional rule that image quality is determined solely by hardware, unleashing the power of “computation” across the entire imaging chain. In 2024, AI’s groundbreaking contributions to fundamental science were recognized with two Nobel Prizes, marking its evolution into a core engine for paradigm shifts in scientific discovery. “AI + computational optical imaging” is now driving a profound revolution, moving beyond back-end algorithmic enhancement toward a deep, source-level “genetic fusion” of AI and optical physics models to build a new-generation intelligent sensing system.

Yet this fusion faces core challenges: achieving efficient, interpretable integration of optical physics priors with complex AI models, striking a balance among data scarcity, computational complexity, and real-time demands, and ultimately bridging the gap from delicate laboratory prototypes to robust, reliable large-scale applications.

On the afternoon of September 9, 2026, the “AI + Computational Optical Imaging” New Opportunities Forum will create a high-efficiency dialogue and collaboration platform, deeply connecting the industrial chain spanning computational imaging algorithms, intelligent optical hardware, 3D vision, AI vision platforms, and computational spectroscopy/hyperspectral imaging. Participants will jointly explore how to overcome the common challenges in large-scale commercialization, together with the latest application advances and technological progress.