Run the following in your shell after activating environment: pip install -r requirements.txt

• The ConvGenVisMo dataset is available in the dataset directory.
• The dataset.jsonl file contains all the conversations with their respective labels; the corresponding images can be found in dataset/images.
• The images generated by Dreamstudio are available in the generated_images_dreamstudio directory.

• To generate ChatGPT summaries for the conversations in the dataset, run chatgpt_summaries.ipynb.
• To generate DreamStudio images for the generated summaries, run dreamstudio_imagegen.ipynb.
• To compute and plot metrics comparing ground truth images with DreamStudio generated images, run compute_metrics.ipynb and plot_metrics.ipynb.

We express our gratitude to the following individuals who generously contributed by sharing their captured photographs with us. We extend our sincerest thanks for their invaluable support.

Dr. Maryam Tavakkoli, Dr. Mahdie Niknezhad, Ms. Zahra Hoseinpoor, Ms. Haniye Hayati, Ms. Niloufar Khoshpasand, Ms. Mahnaz Razavi, Ms. Maryam Nosrati, Ms. Hadis Mostafaviyan, Ms. Samira Najfvand, Ms. Neda Sonboli