Probing the 3D Awareness of Visual Foundation Models
This work addresses the problem of understanding 3D awareness in visual foundation models for researchers in computer vision, but it is incremental as it primarily analyzes existing models without proposing new methods.
The paper investigates whether visual foundation models encode 3D structure and consistent surface representations across views, revealing limitations in current models through experiments with task-specific probes and zero-shot inference.
Recent advances in large-scale pretraining have yielded visual foundation models with strong capabilities. Not only can recent models generalize to arbitrary images for their training task, their intermediate representations are useful for other visual tasks such as detection and segmentation. Given that such models can classify, delineate, and localize objects in 2D, we ask whether they also represent their 3D structure? In this work, we analyze the 3D awareness of visual foundation models. We posit that 3D awareness implies that representations (1) encode the 3D structure of the scene and (2) consistently represent the surface across views. We conduct a series of experiments using task-specific probes and zero-shot inference procedures on frozen features. Our experiments reveal several limitations of the current models. Our code and analysis can be found at https://github.com/mbanani/probe3d.