Full talk · edited transcript

How We Trained a Content Video Encoder Without Labeled Data

Evgeny Astafurov25 May 2024

Evgeny Astafurov, AI VK (previously Dzen), MIPT graduate, instructor of the MLOps course at Phystech.

The topic of training semantic representations for video is not fully explored, so the team faced a number of problems and will explain how they solved them.

Continue reading — it's free

Google or a one-time email code. No password.