EFM3D: A Benchmark for Measuring Progress Towards 3D Egocentric Foundation Models
This work addresses the need for standardized evaluation in developing 3D egocentric foundation models for wearable AI applications, though it is incremental as it builds on existing 2D foundation models.
The paper introduces EFM3D, a benchmark for 3D object detection and surface regression using egocentric sensor data, and proposes Egocentric Voxel Lifting (EVL) as a baseline model that outperforms existing methods on this benchmark.
The advent of wearable computers enables a new source of context for AI that is embedded in egocentric sensor data. This new egocentric data comes equipped with fine-grained 3D location information and thus presents the opportunity for a novel class of spatial foundation models that are rooted in 3D space. To measure progress on what we term Egocentric Foundation Models (EFMs) we establish EFM3D, a benchmark with two core 3D egocentric perception tasks. EFM3D is the first benchmark for 3D object detection and surface regression on high quality annotated egocentric data of Project Aria. We propose Egocentric Voxel Lifting (EVL), a baseline for 3D EFMs. EVL leverages all available egocentric modalities and inherits foundational capabilities from 2D foundation models. This model, trained on a large simulated dataset, outperforms existing methods on the EFM3D benchmark.