**Image Classification of Vehicle Make and Model Using Deep ConvNet**
Rajitha Meka, Rupa Nath, Syed Hasib Akhter Faruqui
*Department of Mechanical Engineering*
*University of Texas at San Antonio, San Antonio, Texas, USA*
{rajithameka14,nathrupa,shafnehal}@gmail.com
**Dataset:** The image data can be found in [http://ai.stanford.edu/%7Ejkrause/cars/car_dataset.html [1]. This directory contains a total of 16,185 images of 196 classes of cars.
**Outcome:**
Our aim is to use a convolutional neural network framework to train and categorize the classes of Make, Model & Year of the cars
**Project Definition:**
We are using a existing Dataset of cars available which have a total of 16185 images belonging to 196 classes of cars. For the purpose of spliting the dataset into training and testing set, each class has been split roughly in a 50-50 split. Thus a total of 8144 testing images and 8041 testing images. It is to be mentioned that the classes are at the level of Make,Model & Year.
Using deep neural network models, LeNet and ConvNet, classification of camera images of cars is studied. The images are taken at different angle and lighting conditions. The background behind the cars are also varient.
A number of information can be extracted from this image data. Like the color of the car, the direction it's facing. The make and model of the car and so on. For this project we are aiming for classifying the make,model & year of the car.
[1]: Jonathan Krause, Michael Stark, Jia Deng, Li Fei-Fei, 3D Object Representations for Fine-Grained Categorization. 4th IEEE Workshop on 3D Representation and Recognition, at ICCV 2013 (3dRR-13). Sydney, Australia. Dec. 8, 2013.
In [ ]: