SimCLR: Cosine Similarity
Beginner
The problem, from first principles
SimCLR (Chen et al., 2020) learns image features by making two augmented views of the same image agree. Agreement is measured by cosine similarity, which ignores the length of the vectors and only compares their direction.
From theory to code
Implement cosine_similarity(a, b), returning the cosine of the angle between two vectors.
Constraints
- Return 0.0 for a zero vector instead of dividing by zero.
Hints
Hint 1
Compute the dot product, divide by the product of the two norms.
Keywords
research-papersunsupervisedcontrastiveself-supervised
Python
Run your code to see results
No results yet
Run your code to check it against the examples, or submit to run every test.