Sources
Explore the references and resources that supported our project.
ADArticho
Datasets
New Yorker Caption Contest (NYCC):
Data sourced from the New Yorker Caption Contest archives, containing captions submitted to the weekly contest.
Oxford Humor in Context (OHIC):
Data sourced from the Oxford University Research Archive, containing approximately 2.9 million humorous image-text pairs.
Tools and Libraries
Python
Used for data preprocessing, analysis, and modeling.
Pandas, NumPy
Libraries for data manipulation and numerical computations.
Matplotlib, Seaborn
Libraries for data visualization.
SciPy
Used for statistical analysis, including correlation computations.
Plotly
Library for creating interactive visualizations.
Transformers
Used for generating embeddings and zero-shot classification.
Sentence Transformers
Used for generating sentence embeddings.
HTML5, CSS3, JS
Technologies used to design and develop this interactive website.
References
- The New Yorker Cartoon Caption Contest Website [1]
- The New Yorker Cartoon Caption Contest Dataset [2]
- Caption Contest Data Examples [3]
- Humor in AI: Massive Scale Crowd-Sourced Preferences and Benchmarks for Cartoon Captioning [4]
- OxfordTVG-HIC: Can Machine Make Humorous Captions from Images? (arXiv preprint) [5]
- OxfordTVG-HIC: Can Machine Make Humorous Captions from Images? (Dataset Website) [6]
Note: The project was built using Artificial Intelligence (Gemini Nano Banana) to generate some images featured on this website.
Explore More
Dive deeper into our datasets, methods, and findings: