BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
X-WR-CALNAME:Cornell Institute for Digital Agriculture (CIDA) Seminar Serie
 s: AI for Scientists: Perception & Discovery - Jennifer Sun
X-WR-TIMEZONE:Eastern Time (US & Canada)
BEGIN:VEVENT
DTSTAMP:20260714T072008Z
UID:tag:localist.com\,2008:EventInstance_44941879626321
DTSTART:20231204T170000Z
DTEND:20231204T180000Z
DESCRIPTION:This seminar is a hybrid event. REGISTRATION REQUIRED for Zoom:
  https://cornell.zoom.us/meeting/register/tJclceCtrDsjHtNPqTkcNT-hXHcgUTfB
 E59O\n\nWith rapidly growing amounts of experimental data\, machine learni
 ng is increasingly crucial for automating scientific data analysis. Howeve
 r\, many real-world workflows demand expert-in-the-loop attention and requ
 ire models that not only interface with data\, but also with experts and d
 omain knowledge. My research develops full stack solutions that enable sci
 entists to scalably extract insights from diverse and messy experimental d
 ata with minimal supervision. My approaches learn from both data and exper
 t knowledge\, while exploiting the right level of domain knowledge for gen
 eralization. In this talk\, I will present progress towards developing aut
 omated scientist-in-the-loop solutions\, including methods that automatica
 lly discover meaningful structure from data such as self-supervised keypoi
 nts from videos of diverse behaving organisms. I will also present methods
  that use these interpretable structures to inject domain knowledge into t
 he learning process\, such as guiding representation learning using symbol
 ic programs of behavioral features computed from keypoints. My aim is to e
 nable AI that collaborates with scientists to accelerate the scientific pr
 ocess.\n\nJennifer is a research scientist at Google\, joining Cornell CS 
 as an assistant professor in August 2024. Her research focuses on computer
  vision and machine learning methods that can be integrated into real-worl
 d workflows involving expert-in-the-loop interactions. She aims to acceler
 ate scientific discovery and optimize expert attention in real-world workf
 lows\, tackling challenges including annotation efficiency\, model interpr
 etability and generalization\, and semantic structure discovery. Jennifer 
 received her PhD from Caltech in September 2023. \n\nThe Cornell Institute
  for Digital Agriculture (CIDA)\, a faculty led initiative focused on crea
 ting a strong voice in the emerging area of Digital Agriculture (DA)\, inv
 ites Jennifer Sun to present her research for CIDA’s monthly seminar ser
 ies.\n\nBackground on the Cornell Institute for Digital Agriculture\n\nAn 
 interdisciplinary group of Cornell University faculty began meeting in ear
 ly 2017 to formulate an Initiative for Digital Agriculture\, believing tha
 t Cornell is uniquely equipped to lead in this emerging arena that will be
 nefit the public for generations. We define DA to mean the application of 
 computational and information technologies coupled with nanotechnology\, b
 iology\, systems engineering and economics to both the research and operat
 ional sides of agriculture and food production. With approximately 100 fac
 ulty from 5 Cornell colleges participating\, we are collaborating with ext
 ernal stakeholders to shape and implement a research agenda for DA that wi
 ll build a pipeline of discovery and innovations for the next 10+ years. P
 lease contact Gabriela Cestero at gc423@cornell.edu with any questions
GEO:42.445068;-76.481117
LOCATION:Bill and Melinda Gates Hall\, 122
SUMMARY:Cornell Institute for Digital Agriculture (CIDA) Seminar Series: AI
  for Scientists: Perception & Discovery - Jennifer Sun
URL;VALUE=URI:https://events.cornell.edu/event/cornell_institute_for_digita
 l_agriculture_cida_seminar_series_jennifer_sun
CATEGORIES:Seminar
CATEGORIES:Special Event
END:VEVENT
END:VCALENDAR