Andrei Kuznetsov
Company: AIRI
Talk
Video Encoding Methods for Multimodal Models
We will discuss the present and future of multimodal architectures based on language models in the task of describing videos and answering questions about them.
Speakers
Invited experts
Konstantin Sobolev
Company: AIRI
Other talks on «ML/AI»
- Watch recording
- Watch recordingTalkHall 1
How to Use AI to Find the Right Frame in Thousands of Videos
Aleksandr Sokolov
Company: GPM Data
- Watch recordingPartner’s talk
VK Calls Voice Quality Metric and How We Got There
Ivan Beskrovnyi
Company: VK / VKontakte
- Watch recording
- Watch recordingTalkHall 2
Noise Cancellation for SaluteJazz. Making Your Conversations More Enjoyable
Yana Lysova
Company: Sber
- Watch recordingTalkHall 2
New Compression Standards and Resilience of JPEG AI to Adversarial Attacks
Dmitry Vatolin
Company: MSU
- Watch recording
- Watch recording
- Watch recording
- Watch recordingTalkHall 2
Vision System for the New Autonomous Platform
Georgiy Nikandrov
Company: Yandex Self-driving Technologies
- Watch recording
- Watch recording
- Watch recording
- Watch recordingTalkHall 1
Multimodal AI, Intelligent Agents — the Future Has Arrived, What's Next?
Eugene Ilyushin
Company: Viasat Tech
