Back to Course
Computer Vision Engineering
Module 4 of 8
4. Face Recognition
1. Siamese Networks
We train a network to measure Similarity. Input: 2 Images. Output: Are they the same person? (Yes/No).
2. Triplet Loss
- Anchor: Image of Person A.
- Positive: Another image of Person A.
- Negative: Image of Person B. Goal: Pull Anchor-Positive close. Push Anchor-Negative far away.