BDD100K: A Diverse Driving Video Database with Scalable Annotation Tooling.

TLDR

Existing driving datasets are limited in visual diversity, and while driving imagery is abundant, annotation remains slow and costly due to inadequate tools. The authors aim to develop a scalable annotation system and release a 100K‑video dataset with diverse labels to advance autonomous driving research. They built a scalable annotation platform that automatically generates comprehensive image labels, enabling the collection of a 100K‑video dataset with varied annotations such as bounding boxes, drivable areas, lane markings, and instance segmentation. The resulting dataset offers extensive geographic, environmental, and weather diversity, helping models generalize to unseen conditions. The dataset is available upon request via the provided URL.

Abstract

Datasets drive vision progress and autonomous driving is a critical vision application, yet existing driving datasets are impoverished in terms of visual content. Driving imagery is becoming plentiful, but annotation is slow and expensive, as annotation tools have not kept pace with the flood of data. Our first contribution is the design and implementation of a scalable annotation system that can provide a comprehensive set of image labels for large-scale driving datasets. Our second contribution is a new driving dataset, facilitated by our tooling, which is an order of magnitude larger than previous efforts, and is comprised of over 100K videos with diverse kinds of annotations including image level tagging, object bounding boxes, drivable areas, lane markings, and full-frame instance segmentation. The dataset possesses geographic, environmental, and weather diversity, which is useful for training models so that they are less likely to be surprised by new conditions. The dataset can be requested at this http URL