A machine learning pipeline for extracting decision-support features from traffic scenes¹

Issue title: Agents in Traffic and Transportation

Guest editors: Ana L.C. Bazzan, Ivana Dusparic, Marin Lujak and Giuseppe Vizzari

Article type: Research Article

Authors: Fraga, Vitor A.^a | Schreiber, Lincoln V.^a | da Silva, Marco Antonio C.^b | Kunst, Rafael^a | Barbosa, Jorge L.V.^a | Ramos, Gabriel de O.^{a; *}

Affiliations: [a] Graduate Program in Applied Computing, Universidade do Vale do Rio dos Sinos, São Leopoldo, RS, Brazil | [b] Institute of Informatics, Universidade Federal do Rio Grande do Sul, Porto Alegre, RS, Brazil

Correspondence: [*] Corresponding author. E-mail: gdoramos@unisinos.br.

Note: [1] A previous version of this work appeared in (In Proceedings of the 12th Workshop on Agents in Traffic and Transportation (ATT-2022) (2022) 194–208).

Abstract: Traffic systems play a key role in modern society. However, these systems are increasingly suffering from problems, such as congestions. A well-known way to efficiently reduce this kind of problem is to perform traffic light control intelligently through reinforcement learning (RL) algorithms. In this context, extracting relevant features from the traffic environment to support decision-making becomes a central concern. Examples of such features include vehicle counting on each queue, identification of vehicles’ origins and destinations, among others. Recently, the advent of deep learning has paved to way to efficient methods for extracting some of the aforementioned features. However, the problem of identifying vehicles and their origins and destinations within an intersection has not been fully addressed in the literature, even though such information has shown to play a role in RL-based traffic signal control. Building against this background, in this work we propose a deep learning pipeline for extracting relevant features from intersections based on traffic scenes. Our pipeline comprises three main steps: (i) a YOLO-based object detector fine-tuned using the UAVDT dataset, (ii) a tracking algorithm to keep track of vehicles along their trajectories, and (iii) an origin-destination identification algorithm. Using this pipeline, it is possible to identify vehicles as well as their origins and destinations within a given intersection. In order to assess our pipeline, we evaluated each of its modules separately as well as the pipeline as a whole. The object detector model obtained 98.2% recall and 79.5% precision, on average. The tracking algorithm obtained a MOTA of 72.6% and a MOTP of 74.4%. Finally, the complete pipeline obtained an average error rate of 3.065% in terms of origin and destination counts.

Keywords: Object detection, object tracking, traffic scenes, intersections, deep learning

DOI: 10.3233/AIC-220317

Journal: AI Communications, vol. 37, no. 2, pp. 189-201, 2024