Unsupervised Vehicle Counting via Multiple Camera Domain Adaptation

TL;DR

This paper introduces a novel unsupervised domain adaptation method for vehicle counting in city images, reducing the need for extensive labeled data across multiple cameras to improve scalability.

Contribution

It presents a new methodology for designing vehicle density estimators that adapt across multiple camera domains with minimal labeled data, enhancing scalability.

Findings

Effective vehicle counting with limited labeled data

Improved scalability across multiple city cameras

Potential reduction in annotation costs

Abstract

Monitoring vehicle flows in cities is crucial to improve the urban environment and quality of life of citizens. Images are the best sensing modality to perceive and assess the flow of vehicles in large areas. Current technologies for vehicle counting in images hinge on large quantities of annotated data, preventing their scalability to city-scale as new cameras are added to the system. This is a recurrent problem when dealing with physical systems and a key research area in Machine Learning and AI. We propose and discuss a new methodology to design image-based vehicle density estimators with few labeled data via multiple camera domain adaptations.