/APL

[2024 AAAI] Object-Aware Adaptive-Positivity Learning for Audio-Visual Question Answering

Primary LanguagePython

Object-aware Adaptive-Positivity Learning for Audio-Visual Question Answering(AAAI'2024) [arXiv]

Authors: Zhangbin Li, Dan Guo, Jinxing Zhou, Jing Zhang, and Meng Wang

Requirements

python3.7 +
pytorch1.7.1
numpy
ast

Usage

  1. Cloning this repo

    git clone https://github.com/zhangbin-ai/APL.git

  2. Getting Started

  • Training

    python train.py \
--batch-size 64 \
--epochs 30 \
--lr 1e-4 \
--gpu 0 \
--checkpoint APL_dir \
--checkpoint_file checkpoint01 \
--save_model_flag True \

  • Testing

    python test.py

  1. Citing Us

    If you find this work useful, please consider citing it.

    @article{li2023object,
 title={Object-aware Adaptive-Positivity Learning for Audio-Visual Question Answering},
 author={Li, Zhangbin and Guo, Dan and Zhou, Jinxing and Zhang, Jing and Wang, Meng},
 journal={arXiv preprint arXiv:2312.12816},
 year={2023}
}

  2. Acknowledgement

    The computation is supported by the HPC Platform of Hefei University of Technology.