MEng student’s work on improving pedestrian safety accepted at a top conference

Important research to improve pedestrian safety, led by the Department of Electronic and Electrical Engineering’s (EEE) third year MEng student Tolga Atun, is to be presented at the prestigious International Conference on Computer Vision Theory and Application (VISAPP 2020). 

Walking is identified as one of the most popular forms of transportation across the world but, according to the World Health Organisation (WHO), is getting ever more dangerous, with an estimated 270K pedestrians per year losing their lives around the world. 

In Summer 2019 Tolga had interned as a Research Associate at RISE (Research Centre of Excellence on Interactive media, Smart systems and Emerging technologies), a Centre in Cyprus that partners with University College London (UCL) to identify ways to improve pedestrian safety through innovative new methods.  

Working under the Biometrics for Smart Human Technologies team, he worked to create a computer vision system that automatically detects pathway obstructions and damages, enabling the safety of pedestrians in the process. The system, employing Machine Learning (ML), has the largest data-set created to date for detecting barriers for pedestrians using wearable camera technologies.

Tolga and his team detailed their research in a conference paper, which has been accepted at the prestigious International Conference on Computer Vision (VISAPP 2020) in Malta, 27 – 29 February. The conference aims to become a major point of contact between researchers, engineers and practitioners on the area of computer vision application systems. 

Tolga’s supervisor Prof. Andreas Demosthenous stated:

Paper abstract and introduction (Full paper to be shared post conference)
Authors: Simoni Panayi, Harris Partaourides, Dr. Zenonas Theodosiou, Professor Andreas Lanitis, Tolga Atun. 

Abstract:
Egocentric vision, which relates to the continuous interpretation of images captured by wearable cameras, is increasingly being utilized in several applications to enhance the quality of citizens’ life, especially for those with visual or motion impairments. The development of sophisticated egocentric computer vision techniques requires automatic analysis of large databases of first-person point of view visual data collected through wearable devices. In this paper, we present our initial findings regarding the use of wearable cameras for enhancing the pedestrians’ safety while walking in city sidewalks. For this purpose, we create a first-person database that entails annotations on common barriers that may put pedestrians in danger. Furthermore, we derive a framework for collecting visual lifelogging data and define 24 different categories of sidewalk barriers. Our dataset consists of 1796 annotated images covering 1969 instances of barriers. The analysis of the dataset by means of object classification algorithms, depict encouraging results for further study.

Introduction:
Walking is the most basic and highly popular form of transportation and it is evident today that it is getting more dangerous. According to the World Health Organization (WHO, 2019), 270K pedestrians per year lose their lives around the world. Contemporary cities have to deal with the various problems caused by the increasing amount of technical barriers and damages that occur on the footpaths which endanger the lives of pedestrians (Sas-Bojarska and Rembeza, 2016). Guaranteeing everyday urban safety has always been a central theme for local authorities, addressing remarkable human, social, and economic aspects. The need for clear paths in urban sidewalks, free of barriers, continuous, and in a well maintained condition is of great importance. Thus, the automatic detection of obstructions and damages can have a positive impact on the sustainability and safety of citizens’ communities. The pedestrian detection (Szarvas et al., 2005) is one of the main research areas as an ultimate aim to develop efficient systems to eliminate deaths in traffic accidents. The safety in roads has attracted a large interest in the last years and a number of studies have been presented for both pedestrians (Nesoff et al., 2018;Wang et al., 2012) and drivers (Timmermans et al., 2019). A study on pothole detection was presented by Prathiba el al. (Prathiba et al., 2015) for the identification of different types of cracks on road pavements. Wang et al. (Wang et al., 2012) developed the WalkSafe, a smartphone application for vehicles recognition to help pedestrians cross safely roads. Jain et al.(Jain and Gruteser, 2017) presented an approach based on smartphone images for recognizing the texture of the surfaces in pedestrians’ routes to be used for safety purposes. A mobile application which uses phone sensors was also presented to enhance the safety of the distracted pedestrians (Tung and Shin, 2018). On the other hand, Maeda et al (Maeda et al., 2018) proposed an approach for the detection of several road damages in smartphones using convolution networks. In the realm of safety, the practicality and efficient use of wearable cameras can effectively help increase the safety of pedestrians. The continuous visual data acquisition can lead to the real-time detection of obstructions, warning the wearers of the potential dangers and alerting the authorities for taking maintenance or corrective actions for ensuring the elimination of dangerous spots for pedestrians. Due to the broad use of deep learning algorithms in analyzing visual lifelogging data, the existence of large annotated datasets are more essential than ever before. Although there are available datasets created by wearable or smartphone cameras refer to road safety, none of them is dedicated specifically for the safety of pedestrians in sidewalks. This work outlines a firstperson database, which can be used for the development of techniques for automatic detection of barriers and other damages that pose safety issues to the pedestrians. In addition, we present initial results on the performance of the dataset in a classification scheme using a well-known deep Convolutional Neural Network (CNNs) as a baseline and elaborate on the promising outcomes. 

Rise Research Centre 
RISE is the Research Centre of Excellence in Cyprus focusing on Interactive media, Smart systems and Emerging technologies aiming to empower knowledge and technology transfer in the region. It is a joint venture between the three public universities of Cyprus - University of Cyprus, Cyprus University of Technology, and, Open University of Cyprus- , the Municipality of Nicosia, and two renowned international partners, the Max Planck Institute for Informatics, Germany, and, the University College London, United Kingdom.