BDD100K is a diverse driving dataset for heterogeneous multitask learning. We construct BDD100K, the largest open driving video dataset with 100K videos and 10 tasks to evaluate the exciting progress of image recognition algorithms on autonomous driving. Each video has 40 seconds and a high resolution. The dataset represents more than 1000 hours of driving experience with more than 100 million frames. The videos comes with GPU/IMU data for trajectory information. The dataset possesses geographic, environmental, and weather diversity, which is useful for training models that are less likely to be surprised by new conditions. The dynamic outdoor scenes and complicated ego-vehicle motion make the perception tasks even more challenging. The tasks on this dataset include image tagging, lane detection, drivable area segmentation, road object detection, semantic segmentation, instance segmentation, multi-object detection tracking, multi-object segmentation tracking, domain adaptation, and imitation learning. This repo contains the toolkit and resources for using BDD100K data
Keywords for this software
References in zbMATH (referenced in 4 articles )
Showing results 1 to 4 of 4.
- Marcos Nieto, Orti Senderos, Oihana Otaegui: Boosting AI applications: Labeling format for complex datasets (2021) not zbMATH
- Kuwajima, Hiroshi; Yasuoka, Hirotoshi; Nakae, Toshihiro: Engineering problems in machine learning systems (2020)
- Holger Caesar, Varun Bankiti, Alex H. Lang, Sourabh Vora, Venice Erin Liong, Qiang Xu, Anush Krishnan, Yu Pan, Giancarlo Baldan, Oscar Beijbom: nuScenes: A multimodal dataset for autonomous driving (2019) arXiv
- Xinyu Huang, Peng Wang, Xinjing Cheng, Dingfu Zhou, Qichuan Geng, Ruigang Yang: The ApolloScape Open Dataset for Autonomous Driving and its Application (2018) arXiv