Unlimited Road-scene Synthetic Annotation (URSA) Dataset

By Matt Angus, Mohamed ElBalkini, Samin Rahman Khan, Ali Harakeh, Oles Andrienko, Cody Reading, Steven L. Waslander, Krzysztof Czarnecki. 2018

23 citations recorded by the source corpus.

In training deep neural networks for semantic segmentation, the main limiting factor is the low amount of ground truth annotation data that is available in currently existing datasets. The limited availability of such data is due to the time cost and human effort required to accurately and consistently label real images on a pixel level. Modern sandbox video game engines provide open world environments where traffic and pedestrians behave in a pseudo-realistic manner. This caters well to the collection of a believable road-scene dataset. Utilizing open-source tools and resources found in single-player modding communities, we provide a method for persistent, ground truth, asset annotation of a game world. By collecting a synthetic dataset containing upwards of 1, 000, 000 images, we demonstrate realtime, on-demand, ground truth data annotation capability of our method. Supplementing this synthetic data to Cityscapes dataset, we show that our data generation method provides qualitative as well as quantitative improvements—for training networks—over previous methods that use video games as surrogate.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Evaluation & Benchmarks

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Evaluation & Benchmarks

Citations in the corpus

Hypersim: A Photorealistic Synthetic Dataset for Holistic Indoor Scene Understanding (2020, methodology): “Interactive Annotation Tools for 3D Scenes Our interactive tool is similar in spirit to existing tools for annotating reconstructed triangle meshes [19, 53, 72] and synthetic images from video games [8, 55].”
ProcSy: Procedural Synthetic Dataset Generation Towards Influence Factor Studies Of Semantic Segmentation Networks (2019, methodology): “proposed a different and more salable method of data generation from GTAV in 2018, and published a dataset with more than 1,000,000 images in ideal weather and lighting conditions [1].”
Precise Synthetic Image and LiDAR (PreSIL) Dataset for Autonomous Vehicle Perception (2019, methodology): “The URSA dataset [6] increases the quality of semantic”