Extended Data Table 1 Quantitative comparison of our method against state-of-the-art

From: Low-latency automotive vision with event cameras

  1. (a) Comparison against asynchronous methods in terms of computational complexity, task performance, and energy consumption. Results of event-based methods on the Gen1 detection dataset41 and N-Caltech10142. (b) Comparison of event and image-based detectors on DSEC-Detection. Here, methods are tasked to predict labels 50 ms after the first image, given a single image, and events. Speed refers to the average runtime of event insertion into the GNN averaged over the dataset.
Back to article page