AudioCaps: Generating Captions for Audios in The Wild

This repository contains the code and the dataset for our NAACL-HLT 2019 paper.

• Chris Dongjoo Kim, Byeongchang Kim, Hyunmin Lee, and Gunhee Kim. AudioCaps: Generating Captions for Audios in The Wild. In NAACL-HLT, 2019. (Oral)

The Audio Captioning Task

For a live demo visit our website, https://audiocaps.github.io/

Citation

The code and the dataset are free to use for academic purposes only. If you use any of the material in this repository as part of your work, we ask you to cite:

@inproceedings{kim-NAACL-HLT-2019,
    author    = {Chris Dongjoo Kim and Byeongchang Kim and Hyunmin Lee and Gunhee Kim},
    title     = "{AudioCaps: Generating Captions for Audios in The Wild}"
    booktitle = {NAACL-HLT},
    year      = 2019
}

Last edit: May 30, 2019