BEGIN:VCALENDAR
VERSION:2.0
PRODID:IEEE vTools.Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
BEGIN:DAYLIGHT
DTSTART:20250309T030000
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
RRULE:FREQ=YEARLY;BYDAY=2SU;BYMONTH=3
TZNAME:PDT
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251102T010000
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
RRULE:FREQ=YEARLY;BYDAY=1SU;BYMONTH=11
TZNAME:PST
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250627T043436Z
UID:80F0639E-D37D-4546-9CF4-E49C3C146173
DTSTART;TZID=America/Los_Angeles:20250626T120000
DTEND;TZID=America/Los_Angeles:20250626T130000
DESCRIPTION:This talk addresses modern approaches to image and video compre
 ssion through the lens of energy-efficient hardware design. Traditional co
 decs like JPEG and H.264 are increasingly being challenged by learned comp
 ression techniques based on deep neural networks\, particularly autoencode
 rs. While these methods offer state-of-the-art performance in rate-distort
 ion trade-offs\, their deployment in real-world systems depends critically
  on efficient circuit and architectural design. We will explore the struct
 ure and training of neural compression models\, including variational auto
 encoders and entropy bottlenecks\, followed by the challenges of implement
 ing these models in energy- and area-constrained environments such as mobi
 le devices\, cameras\, and edge computing systems.\n\nSpeaker(s): Mateus G
 rellert\n\nVirtual: https://events.vtools.ieee.org/m/487438
LOCATION:Virtual: https://events.vtools.ieee.org/m/487438
ORGANIZER:vickyhlu@ieee.org
SEQUENCE:65
SUMMARY:IEEE SPS SCV - Energy-Efficient Neural Image and Video Compression 
URL;VALUE=URI:https://events.vtools.ieee.org/m/487438
X-ALT-DESC:Description: <br /><p class="MsoNormal"><span lang="EN-AU">This 
 talk addresses modern approaches to image and video compression through th
 e lens of energy-efficient hardware design. Traditional codecs like JPEG a
 nd H.264 are increasingly being challenged by learned compression techniqu
 es based on deep neural networks\, particularly autoencoders. While these 
 methods offer state-of-the-art performance in rate-distortion trade-offs\,
  their deployment in real-world systems depends critically on efficient ci
 rcuit and architectural design. We will explore the structure and training
  of neural compression models\, including variational autoencoders and ent
 ropy bottlenecks\, followed by the challenges of implementing these models
  in energy- and area-constrained environments such as mobile devices\, cam
 eras\, and edge computing systems.</span></p>
END:VEVENT
END:VCALENDAR