On 1/7 I will be talking about Siamese networks (short lacture in a meet up)
I will be talking about my last 2 (or three) blog posts around Siamese networks:
1.Basic post about How Siamese networks play with the semi supervised domain, when do they make sense, and more importantly how do we make them fail miserably.
2.Second, more advanced post about what can we reasonably expect from them.
In this meetup. we will get to know Siamese networks a little bit better, Starting from definition, what can we theoretically expect from them, and what actually seems to work. We will mention triplet loss in a word, and try to talk about further research directions if we will have the time. Don’t come expecting to become an expert, do come expecting to get a cool new direction of thought about semi supervised learning