BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//bbuzz22//speaker//PUPNAC
BEGIN:VTIMEZONE
TZID:Europe/Berlin
BEGIN:DAYLIGHT
DTSTART:20210614T000000
TZNAME:CEST
TZOFFSETFROM:+0200
TZOFFSETTO:+0200
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20211031T030000
RDATE:20221030T030000
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20220327T030000
RDATE:20230326T030000
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
SUMMARY:Muves: Multimodal and multilingual vector search with Hardware Acc
 eleration - Aarne Talman\, Dmitry Kan
DTSTART;TZID=Europe/Berlin:20220614T160000
DTEND;TZID=Europe/Berlin:20220614T164000
DTSTAMP:20261112T193019Z
UID:pretalx-bbuzz22-YFLCEZ@pretalx.com
DESCRIPTION:Bringing multimodal experience into search journey became of h
 igh interest lately: searching images with text\, or looking inside an aud
 io file\, combining that with the rgb frames of a video stream. Today\, ve
 ctor search algorithms (like FAISS\, HNSW\, BuddyPQ) and databases (Vespa
 \, Weaviate\, Milvus and others) make these experiences a reality. But wha
 t if you as a user would like to stay with the familiar Elasticsearch / Op
 enSearch AND leverage the vector search at scale? In this talk we will tak
 e a hardware acceleration route to build a vector search experience over p
 roducts and will show how you can blend the worlds of neural search with s
 ymbolic filters. \n\nWe will discuss use cases where adding multimodal and
  multilingual vector search will improve recall and compare results from E
 lasticsearch/OpenSearch with and without the vector search component using
  tools like Quepid. We will also investigate different fine-tuning approac
 hes and compare their impact on different quality metrics.\n\nWe will demo
 nstrate our findings using our end-to-end search solution Muves which comb
 ines traditional symbolic search with multimodal and multilingual vector s
 earch and includes an integrated fine-tuner for easy domain adaptation of 
 pre-trained vector models.\n\nThe Search track is presented by OpenSource 
 Connections
LOCATION:Frannz Salon
URL:https://pretalx.com/bbuzz22/talk/YFLCEZ/
END:VEVENT
END:VCALENDAR
