Summary of Seeing Faces in Things: a Model and Dataset For Pareidolia, by Mark Hamilton et al.
Seeing Faces in Things: A Model and Dataset for Pareidolia
by Mark Hamilton, Simon Stent, Vasha DuTell, Anne Harrington, Jennifer Corbett, Ruth Rosenholtz, William T. Freeman
First submitted to arxiv on: 24 Sep 2024
Categories
- Main: Computer Vision and Pattern Recognition (cs.CV)
- Secondary: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Information Retrieval (cs.IR); Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper investigates face pareidolia, the tendency to perceive faces in random or non-face stimuli, from a computer vision perspective. Researchers presented an image dataset called “Faces in Things,” comprising 5,000 web images with human-annotated pareidolic faces. They examined how state-of-the-art human face detectors perform on this task and found significant differences between humans and machines. The study suggests that the evolutionary need to detect animal and human faces may explain these differences. Additionally, a simple statistical model of pareidolia was proposed, which predicts when certain image conditions are most likely to induce pareidolia. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how computers can be tricked into seeing faces where there aren’t any. This is called face pareidolia and happens when we see faces in unexpected places like coffee stains or clouds. The researchers created a big dataset of images with human-annotated faces that might not actually be faces. They tested how good computers are at finding these fake faces and found that they’re much worse than humans. This makes sense because our brains have evolved to find faces in order to survive, but it’s still an interesting difference between us and machines. |
Keywords
» Artificial intelligence » Statistical model
