Vehicle Type Image Dataset (Version 2): VTID2
Published: 24 November 2021| Version 1 | DOI: 10.17632/htsngg9tpc.1
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