One Thousand and One Hours: Self-driving Motion Prediction Dataset

TLDR

The paper introduces the largest self‑driving motion‑prediction dataset, comprising over 1,000 hours of data, to advance machine‑learning systems. The dataset was collected by 20 autonomous vehicles on a fixed Palo Alto route over four months, yielding 170,000 25‑second scenes with perception outputs and a high‑definition semantic map and aerial view. Using this dataset markedly improves performance on core self‑driving tasks, and together with the software kit it is the most detailed resource for motion forecasting, planning, and simulation. The dataset is available at http://level5.lyft.com/.

Abstract

Motivated by the impact of large-scale datasets on ML systems we present the largest self-driving dataset for motion prediction to date, containing over 1,000 hours of data. This was collected by a fleet of 20 autonomous vehicles along a fixed route in Palo Alto, California, over a four-month period. It consists of 170,000 scenes, where each scene is 25 seconds long and captures the perception output of the self-driving system, which encodes the precise positions and motions of nearby vehicles, cyclists, and pedestrians over time. On top of this, the dataset contains a high-definition semantic map with 15,242 labelled elements and a high-definition aerial view over the area. We show that using a dataset of this size dramatically improves performance for key self-driving problems. Combined with the provided software kit, this collection forms the largest and most detailed dataset to date for the development of self-driving machine learning tasks, such as motion forecasting, motion planning and simulation. The full dataset is available at http://level5.lyft.com/.