Automatic Ground Truths: Projected Image Annotations for Omnidirectional Vision

TL;DR

This paper introduces a new omnidirectional video dataset with automatically annotated object positions, facilitating training and evaluation of scene understanding algorithms in spherical imagery.

Contribution

It provides a novel dataset with automatically generated ground truth annotations for omnidirectional vision, along with calibration tools and error estimation methods.

Findings

Dataset enables improved training of object detection algorithms

Automated annotations reduce manual labeling effort

Software tools facilitate calibration and comparison

Abstract

We present a novel data set made up of omnidirectional video of multiple objects whose centroid positions are annotated automatically. Omnidirectional vision is an active field of research focused on the use of spherical imagery in video analysis and scene understanding, involving tasks such as object detection, tracking and recognition. Our goal is to provide a large and consistently annotated video data set that can be used to train and evaluate new algorithms for these tasks. Here we describe the experimental setup and software environment used to capture and map the 3D ground truth positions of multiple objects into the image. Furthermore, we estimate the expected systematic error on the mapped positions. In addition to final data products, we release publicly the software tools and raw data necessary to re-calibrate the camera and/or redo this mapping. The software also provides a simple framework for comparing the results of standard image annotation tools or visual tracking systems against our mapped ground truth annotations.