Object Recognition from very few Training Examples for Enhancing Bicycle Maps

TL;DR

This paper presents a novel object recognition system for cycling infrastructure that effectively learns from only 15 examples per class by combining CNNs and random forests, enabling accurate detection and mapping of traffic signs with minimal labeled data.

Contribution

It introduces a method that reduces the need for large labeled datasets by integrating CNNs and random forests, tailored for recognizing traffic signs with very few training examples.

Findings

Achieves accurate traffic sign recognition with only 15 examples per class.

Significantly accelerates image processing through a fully convolutional network.

Outperforms existing methods like Faster R-CNN in low-data scenarios.

Abstract

In recent years, data-driven methods have shown great success for extracting information about the infrastructure in urban areas. These algorithms are usually trained on large datasets consisting of thousands or millions of labeled training examples. While large datasets have been published regarding cars, for cyclists very few labeled data is available although appearance, point of view, and positioning of even relevant objects differ. Unfortunately, labeling data is costly and requires a huge amount of work. In this paper, we thus address the problem of learning with very few labels. The aim is to recognize particular traffic signs in crowdsourced data to collect information which is of interest to cyclists. We propose a system for object recognition that is trained with only 15 examples per class on average. To achieve this, we combine the advantages of convolutional neural networks and random forests to learn a patch-wise classifier. In the next step, we map the random forest to a neural network and transform the classifier to a fully convolutional network. Thereby, the processing of full images is significantly accelerated and bounding boxes can be predicted. Finally, we integrate data of the Global Positioning System (GPS) to localize the predictions on the map. In comparison to Faster R-CNN and other networks for object recognition or algorithms for transfer learning, we considerably reduce the required amount of labeled data. We demonstrate good performance on the recognition of traffic signs for cyclists as well as their localization in maps.