EventHub: Data Factory for Generalizable Event-Based Stereo Networks without Active Sensors
2026-04-02 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors introduce EventHub, a new way to train deep learning models that use event cameras for stereo vision without needing expensive ground truth data from active sensors. Instead, they create helpful training data from regular color images using advanced image synthesis or paired event data. By using this approach, they adapt existing stereo models designed for RGB images to work well with event data, achieving better generalization. Their tests on common event stereo datasets show that this method improves performance, even helping RGB stereo models work better in hard conditions like nighttime.
Event camerasStereo visionDeep learningNovel view synthesisProxy annotationsGeneralizationRGB imagesData distillationActive sensorsNighttime imaging
Authors
Luca Bartolomei, Fabio Tosi, Matteo Poggi, Stefano Mattoccia, Guillermo Gallego
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.