Connectionists: Call for Papers: Special Session on Multimodal and Lifelong Learning at ICAART

[Wiering, M.A.]

Dear colleagues,


Hereby you are invited to submit a paper to a special session on Multimodal
and Lifelong Learning (MALM) organized at ICAART 2019 in Prague, see:
http://www.icaart.org/MALM.aspx

Modern machine learning has proven to be very successful in unimodal
applications (image OR text OR sound). Furthermore, the successes are
usually based on idealized training conditions with nicely packaged
benchmark tests. The reality of historical-document retrieval is quite
different. Optical character recognition is too limited to handle the
variety of visual patterns in such image collections: text, graphics,
tabular structures, doodles and diverse image problems make this a
challenging playing field. Such systems start with zero labels, the labels
change over time and the data is neither stationary nor ergodic. This
session is intended for researchers who have picked up challenges in
multimodal machine learning, possibly even in a time-varying context.

Topics:

- Multimodal Deep Learning
- Image and Text Correspondence
- Data-driven Semantics
- Deep Multimodal Semantic Embeddings
- Historical Documents
- Interactive Data Mining

Organizers:
Prof. dr. Lambert Schomaker, University of Groningen
Dr. Marco Wiering, University of Groningen

IMPORTANT DATES

Paper Submission: December 20, 2018
Authors Notification: January 7, 2019
Camera Ready and Registration: January 15, 2019
Conference: February 19-21, 2019

Submitted papers should be between 6 and 12 pages long. Please note that a
double-blind reviewing system is used. For more information about paper
guidelines, please see:
http://www.icaart.org/CallForPapers.aspx
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20181128/7f7bef4c/attachment.html>