Vehicle Type Image Dataset (Version 2): VTID2

Published: 25 November 2021| Version 2 | DOI: 10.17632/htsngg9tpc.2
Contributors:
,

Description

After creating VTID, the researchers decided to extend the collection process to create another larger dataset to add further diversity to the dataset in order to avoid data overfitting. Finally, the new dataset, called "Vehicle Type Image Dataset 2 (VTID2)", consisted of 4,356 image samples that could be separated into five vehicle type classes as follows: 1,230 sedans, 1,240 pick-ups, 680 SUVs, 606 hatchbacks, and 600 other vehicle images.

Files

Institutions

Mahasarakham University

Categories

Image Classification, Convolutional Neural Network

License