Bag Accessories Detection Dataset

Published: 30 July 2024| Version 1 | DOI: 10.17632/tzmf2ztdwp.1
Contributors:
,
, Sohan Chavan, Bhaveshkumar Vagadiya, Debayan Talapatra

Description

The "Bag Accessories Detection" dataset is meticulously curated to enhance the object detection model, specifically focusing on identifying various accessories of bags such as handles and straps. Comprising a total of 2,528 images, this dataset captures a diverse array of bags including hard bags, soft bags, backpacks, long bags, and duffle bags. The images were collected between October and December 2023, primarily from the India Innovation Centre Vanderlande in Pune, as well as from platforms like Roboflow, YouTube, iStock, and Shutterstock. Each image is captured using an iPhone 13 12MP rear camera, ensuring high resolution and clarity, and is taken from multiple angles and distances to simulate real-world scenarios. The images are labeled with three distinct classes: "bag," "handle," and "straps," providing a detailed annotation that will facilitate the precise training of object detection models. This dataset is particularly valuable for the airport automation industry, aimed at improving the efficiency and accuracy of automated baggage handling systems.

Files

Categories

Machine Learning, Computer Engineering, Computer Imaging, Deep Learning, Image Analysis

Funding

Vanderlande Industries Pvt. Ltd. Pune

Licence