How would you solve a machine learning vision problem? (Specific details were given to this problem)
Sigiloso
I described an in detail architecture for image embeddings. The conversation had many follow up questions, and was quite fun and interesting.