Papers with adaptive attention-based model
BAN-Cap: A Multi-Purpose English-Bangla Image Descriptions Dataset (2022.lrec-1)
Copied to clipboard
| Challenge: | BAN-Cap dataset is based on the widely used Flickr8k dataset, which is used to collect captions of images from qualified annotators. |
| Approach: | They propose to use a dataset to collect Bangla captions from qualified annotators and to evaluate the models for the task. |
| Outcome: | The proposed model outperforms state-of-the-art models for Bangla captioning and English-Bangla translation. |