Search papers, labs, and topics across Lattice.
This study introduces ImagingBench, a comprehensive benchmark assessing the capabilities of vision-language models (VLMs) and agentic AI across 20 computational imaging tasks, including ray optics and inverse reconstruction. The findings reveal that while these models perform well on semantic visual tasks, they significantly lag behind specialized methods in handling complex computational imaging challenges, particularly in areas like lensless imaging and holography. Notably, the research highlights that despite generating visually plausible outputs, the fidelity of these models remains inadequate, underscoring a critical gap between semantic understanding and physical imaging competence.
Agentic AI struggles with computational imaging tasks, revealing a stark divide between visual plausibility and physical fidelity.
Vision-language models (VLMs) and agentic AI have shown strong performance on semantic visual tasks, but it remains unclear whether they can handle the physics and inverse problems that underlie computational imaging. We present ImagingBench, a benchmark of 20 computational imaging tasks spanning five categories: ray and wave optics, image signal processing, inverse reconstruction, computational sensing, and calibration. ImagingBench evaluates three complementary settings: Expert, fixed expert-guided inverse reconstruction; Planner, planner-guided inverse reconstruction; and Forward, forward-system simulation for consistency checking. We benchmark leading proprietary and open-source image-centric multimodal systems, including Gemini, GPT, and Qwen, and compare them with representative task-specific non-agentic baselines. Across tasks, agentic models remain consistently weaker than specialized methods, especially on computational sensing problems such as lensless imaging, event-based reconstruction, time-of-flight imaging, and holography. Planner guidance provides only modest and inconsistent gains over the fixed-prompt Expert baseline. Although the models often generate visually plausible outputs, their reference-based fidelity remains poor, revealing a substantial gap between semantic visual competence and physically grounded imaging performance. ImagingBench provides a unified testbed for measuring this gap and tracking progress in agentic AI for computational imaging.
