BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.4.0.1//EN
TZID:Europe/Paris
X-WR-TIMEZONE:Europe/Paris
BEGIN:VEVENT
UID:406@isdm.umontpellier.fr
DTSTART;TZID=Europe/Paris:20241017T140000
DTEND;TZID=Europe/Paris:20241017T140000
DTSTAMP:20260828T115549Z
URL:https://isdm.umontpellier.fr/events/some-challenges-around-retraining-
 generative-models-on-their-own-data-2/
SUMMARY:Some Challenges Around Retraining Generative Models on their Own Da
 ta
DESCRIPTION:Room 02.124\, Building 5\, St Priest campus\n\nMachine Learning
  in Montpellier\, Theory &amp\; Practice\n\nDeep generative models have ma
 de tremendous progress in modeling complex data\, often exhibiting generat
 ion quality that surpasses a typical human&#x27\;s ability to discern the 
 authenticity of samples. Undeniably\, a key driver of this success is enab
 led by the massive amounts of web-scale data consumed by these models. Due
  to these models&#x27\; striking performance and ease of availability\, th
 e web will inevitably be increasingly populated with synthetic content. Su
 ch a fact directly implies that future iterations of generative models wil
 l be trained on both clean and artificially generated data from past model
 s. In addition\, in practice\, synthetic data is often subject to human fe
 edback and curated by users before being used and uploaded online. For ins
 tance\, many interfaces of popular text-to-image generative models\, such 
 as Stable Diffusion or Midjourney\, produce several variations of an image
  for a given query which can eventually be curated by the users. In this t
 alk we will discuss the impact of training generative models on mixed data
 sets---from classical training on real data to self-consuming generative m
 odels trained on purely synthetic curated data.\n\nMachine Learning in Mon
 tpellier\, Theory &amp\; Practice
ATTACH;FMTTYPE=image/jpeg:https://isdm.umontpellier.fr/wp-content/uploads/
 2026/06/ml-mtp-gC78d5.png
CATEGORIES:ML MTP
LOCATION:Room 02.124 Building 5\, St Priest Campus\, Montpellier\, 34000\, 
 France
X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=St Priest Campus\, Montpell
 ier\, 34000\, France;X-APPLE-RADIUS=100;X-TITLE=Room 02.124 Building 5:geo
 :0,0
END:VEVENT
BEGIN:VTIMEZONE
TZID:Europe/Paris
X-LIC-LOCATION:Europe/Paris
BEGIN:DAYLIGHT
DTSTART:20240331T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR