The EuroCity Persons Dataset: A Novel Benchmark for Object Detection

TL;DR

The paper introduces the EuroCity Persons dataset, a large, diverse, and detailed benchmark for pedestrian and cyclist detection in urban traffic scenes, enabling improved deep learning model evaluation.

Contribution

It presents a new extensive dataset with high-quality annotations collected across Europe, and evaluates multiple state-of-the-art detectors to establish baseline performances.

Findings

EuroCity Persons dataset contains over 238,200 person instances in 47,300 images.

Training on this dataset improves generalization of detectors across different conditions.

Dataset diversity and annotation quality significantly affect detection performance.

Abstract

Big data has had a great share in the success of deep learning in computer vision. Recent works suggest that there is significant further potential to increase object detection performance by utilizing even bigger datasets. In this paper, we introduce the EuroCity Persons dataset, which provides a large number of highly diverse, accurate and detailed annotations of pedestrians, cyclists and other riders in urban traffic scenes. The images for this dataset were collected on-board a moving vehicle in 31 cities of 12 European countries. With over 238200 person instances manually labeled in over 47300 images, EuroCity Persons is nearly one order of magnitude larger than person datasets used previously for benchmarking. The dataset furthermore contains a large number of person orientation annotations (over 211200). We optimize four state-of-the-art deep learning approaches (Faster R-CNN, R-FCN, SSD and YOLOv3) to serve as baselines for the new object detection benchmark. In experiments with previous datasets we analyze the generalization capabilities of these detectors when trained with the new dataset. We furthermore study the effect of the training set size, the dataset diversity (day- vs. night-time, geographical region), the dataset detail (i.e. availability of object orientation information) and the annotation quality on the detector performance. Finally, we analyze error sources and discuss the road ahead.