BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//CERN//INDICO//EN
BEGIN:VEVENT
SUMMARY:Deep Upstream : Hardware Agnostic GStreamer Analytics
DTSTART;VALUE=DATE-TIME:20230925T104000Z
DTEND;VALUE=DATE-TIME:20230925T112000Z
DTSTAMP;VALUE=DATE-TIME:20260718T002743Z
UID:indico-contribution-258@indico.freedesktop.org
DESCRIPTION:Speakers: Daniel Morin (Collabora Inc.)\nWith the growing powe
 r of machine-learning\, the time has come for GStreamer to support complex
 \, platform-independent analytic pipelines for tracking\, super-resolution
 \, noise filtering\, speech recognition and more general analysis of timed
  data streams. We discuss a new flexible and efficient design to address t
 hese problems\, without vendor or framework lock-in\, which can easily int
 eroperate with existing downstream approaches.\n \nTo achieve this goal\, 
 we have designed new framework-independent graph-based infrastructure usin
 g the existing GstMeta structure to store complex metadata and their relat
 ionships. We have also generalized the existing ONNX-based object detector
  to easily support many new inference models targeting a variety of hardwa
 re backends\, and have built a new OSD to visualize the generated analytic
 s metadata. Care has been taken to ensure efficient pipelines with support
  for batch processing and zero-copy. Finally\, we have built a bridge to n
 on-GStreamer land with a new cloud metadata sink that can send analytics r
 esults to cloud servers.\n\nWe will also present a demo at the end of the 
 talk showcasing a complex two-phase video analysis pipeline.\n\nhttps://in
 dico.freedesktop.org/event/5/contributions/258/
LOCATION:Palexco Room 2
URL:https://indico.freedesktop.org/event/5/contributions/258/
END:VEVENT
END:VCALENDAR