Vehicle Attribute Recognition by Appearance: Computer Vision Methods for Vehicle Type, Make and Model Classification

TL;DR

This paper reviews computer vision techniques for recognizing vehicle attributes like type, make, and model from images, comparing classification and metric learning methods in a simulated real-world setting.

Contribution

It provides a comprehensive survey of vehicle attribute recognition algorithms and compares two different approaches through experimental evaluation.

Findings

Classification and metric learning methods show different strengths.

Simulated real-world scenario highlights practical challenges.

Experimental results inform future vehicle recognition system design.

Abstract

This paper studies vehicle attribute recognition by appearance. In the literature, image-based target recognition has been extensively investigated in many use cases, such as facial recognition, but less so in the field of vehicle attribute recognition. We survey a number of algorithms that identify vehicle properties ranging from coarse-grained level (vehicle type) to fine-grained level (vehicle make and model). Moreover, we discuss two alternative approaches for these tasks, including straightforward classification and a more flexible metric learning method. Furthermore, we design a simulated real-world scenario for vehicle attribute recognition and present an experimental comparison of the two approaches.