Science & AI
How our LLM Engine works, epigenetics, DNA test vs facial AI comparison, tips for the perfect photo. The tech & advanced science side.
Artificial intelligence has transformed resemblance measurement in recent years. Where you used to need an expert's eye to compare two faces, a deep learning model like ArcFace now produces a reproducible numerical score in less than a second. This category demystifies these technologies — facial embeddings, feature vectors, convolutional neural networks — without unnecessary jargon.
It also compares competing approaches: DNA tests (useful for genealogy, oversized for perceived resemblance), naked-eye comparison (very subjective, prone to cognitive bias), static photo vs video (photo wins in 95% of cases). You will find tips for optimising your photo quality before a test, and understanding the AI's precision limits — especially on newborns <6 months where faces are still little differentiated.
To go further, the glossary details each technical term cited in these articles.
4 articles
How Does Our Facial Recognition AI Work?
Behind the score, our proprietary LLM Engine analyses dozens of facial landmarks in seconds. Here's how.
Epigenetics: Can the Environment Change Resemblance?
When environment (diet, stress, exposure) modifies gene expression — and therefore resemblance.
DNA Test vs Facial AI: Which to Choose?
Honest comparison on price, accuracy, time, privacy — when to choose one or the other.
How to Take the Perfect Photo for a Test
7 simple rules for genuinely accurate results — and 3 traps to absolutely avoid.
🪞 You are reading about resemblance — try it yourself
Start my free test →