Skip to main content

Datasets

Standard Dataset

Dataset for Image Captioning System (in Bangla)

Citation Author(s):
Toshiba Kamruzzaman
Submitted by:
Toshiba Kamruzzaman
Last updated:
DOI:
10.21227/4tsj-yn92
Data Format:
No Ratings Yet

Abstract

We have developed this dataset for the Bangla image caption. Here, we have recorded  500 images with one caption of each. Basically the lifestyle, festivals are mainly focused in this dataset. We have accomplished rice/harvest festivals, snake charming, palanquin, merry-go-round, slum, blacksmith, potter, fisherman, tat shilpo, jamdani, shutki chash, date juice, hal chash, tokai, pohela falgun, gaye holud, etc.

   

Instructions:

Content

This is an image Dataset containing 500 images with captions that a human may give to describe the image i.e captioning the image. Here, the lifestyle and important events of Bangladesh are mainly highlighted.

File Information

The Dataset contains two directory -

  • Images - This directory contains 500 Images.

  • Text - This file contains a caption for each Image present in the Images Directory.