EventHub: A Data Factory for Generalizable Event-Based Stereo Networks

Forum topic · 小凯 · 2026-04-04

Summary

EventHub is a novel framework proposed by Luca Bartolomei, Fabio Tosi, and Matteo Poggi (arXiv:2504.01265, April 2025) for training deep event-based stereo networks without ground-truth annotations from expensive active sensors. Instead, it relies on standard color images: through state-of-the-art novel view synthesis, it generates proxy annotations and proxy events, or only proxy annotations when images are already paired with event data. Acting as a data factory, EventHub enables repurposing state-of-the-art RGB stereo models to process event data, yielding event stereo models with unprecedented generalization capabilities. Experiments on widely used event stereo datasets validate the approach, and the same data distillation mechanism also improves the accuracy of RGB stereo foundation models under challenging conditions such as night scenes.

Overview

Field: Computer Vision Authors: Luca Bartolomei, Fabio Tosi, Matteo Poggi Published: 2025-04-01 arXiv: 2504.01265

Abstract

We propose EventHub, a novel framework for training deep-event stereo networks without ground truth annotations from costly active sensors, relying instead on standard color images. From these images, we derive either proxy annotations and proxy events through state-of-the-art novel view synthesis techniques, or simply proxy annotations when images are already paired with event data.

Using the training set generated by our data factory, we repurpose state-of-the-art stereo models from RGB literature to process event data, obtaining new event stereo models with unprecedented generalization capabilities. Experiments on widely used event stereo datasets support the effectiveness of EventHub and show how the same data distillation mechanism can improve the accuracy of RGB stereo foundation models in challenging conditions such as nighttime scenes.

Key Ideas

No expensive active sensors needed: training data is derived from standard color images instead of ground-truth labels captured with costly depth sensors.
Proxy supervision: novel view synthesis techniques generate proxy annotations and proxy events; when event data is already paired with images, only proxy annotations are needed.
Cross-domain transfer: state-of-the-art RGB stereo models can be repurposed for event-based stereo, achieving strong generalization.
Bonus: the same distillation mechanism also improves RGB stereo foundation models in challenging conditions like night scenes.