Archives
Audio Reactivity
.
2024
Touch Designer
.
Touch Designer
https://drive.google.com/file/d/1s0tgDF48j47uywVpKAp2FgKhwmA34VJJ/view?usp=sharing
DOGGLEGÄNGER
.
2026
Web Design . API . AI
Dogglegänger uses facial recognition and machine learning algorithms that analyze facial features and patterns. By comparing a user's photo with a vast database of dog images, the system identifies potential matches and suggests the most likely canine match based purely on physical features and expressions.
HTML . CSS. Javascript . Figma
Turns out it's not just a theory but an actual phenomenon according to National Geographic, BBC, American Kennel Club, USA Today, Business Insider, New York Post, CNN, The Guardian, Discover Magazine, The Economist, Monaco Tribune, Reader's Digest, etc. Most of the studies focus on personality - shared energy, lifestyle, even stress levels bleeding between owner and dog over time. But what if it's pushed further? What if the resemblance isn't just behavioural, but actually visible? Dogglegänger uses facial recognition and machine learning algorithms that analyse facial features and patterns. By comparing a user's photo with a vast database of dog images, the system identifies potential matches and suggests the most likely canine match based purely on physical features and expressions.
The matching works by comparing facial features between the uploaded photo and a database of 3,853 dogs. Each feature is measured by a weighted comparison, where the closer the proportions, the closer the match. The system then finds the dog with the closest overall match. To determine the best calculation method, I used AI to arrive at three approaches: Z-Score Normalisation, Cosine Similarity, and Weighted Euclidean Distance, and then analysed the results to arrive at my final code.
King Midas and the Golden Touch
.
.
.
.
