MultiModFuseNeT

The BMCC corpus link: https://data.mendeley.com/datasets/htpk9y2pwf/2 Article Link: https://www.sciencedirect.com/science/article/pii/S095070512501130X?dgcid=coauthor

Installation

Requirements

• Python 3.8+
• PyTorch 1.12+
• Transformers 4.25+
• CUDA-capable GPU (recommended)
• OpenCV (cv2)
• Pandas, NumPy, scikit-learn
• CLIP (for image-text similarity scoring)

Setup

1. Clone the repository:

git clone https://github.com/mrhossain/MultiModFuseNeT
cd MultiModFuseNeT

2. Create and activate a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows, use: venv\Scripts\activate

3. Install dependencies:

pip install torch torchvision
pip install transformers
pip install pandas numpy scikit-learn opencv-python
pip install clip
pip install openai  # For GPT-based experiments