Explainable Real Time Vehicle Make Model Recognition Using Bag of Expressions Features

Vehicle Make and Model Recognition (VMMR) has been an area of research in last few years. research in this topic will improve transportation systems.  This has application or can be applied in traffic surveillance and traffic management. There is also increasing demand for traffic monitoring solutions.  Many of the vehicle identification techniques are based on manual observation, license plate recognition. These system does not perform up to satisfactory level in the abnormal weather conditions. This faces problem in low light , occlusion, camera position with respect to vehicle, motion blur, and poor image quality. [1]

This review paper presents a comprehensive study of existing Vehicle Make and Model Recognition approaches, datasets, feature extraction methods, and classification techniques. The paper discusses traditional machine learning methods and deep neural network learning approaches, including Convolutional Neural Networks (CNNs), transfer learning, attention mechanisms, and hybrid frameworks. [1]

The study of the recently published papers shows that deep learning particularly CNN, transformer based architectures, provide better accuracy compared to conventional techniques. Integration of VMMR with edge computing and intelligent surveillance systems enables faster and more scalable deployment.

This topic has many of the applications like traffic monitoring, toll collection, law enforcement, stolen vehicle detection, surveillance, etc. Once a better model is developed many new applications can be identified in coming years. It has difficulty of Introduction of new models, Improvement of existing models.  Which make an big hurdle. Weather conditions and Terrin are not stable in india there is need for consistent improvement.