[visionlist] [CfP] WeaSuL - Workshop on Weakly Supervised Learning @ ICLR 2021

Michael A. Hedderich mhedderich at lsv.uni-saarland.de
Thu Jan 7 10:04:02 -04 2021

Call for Papers


WeaSuL: Workshop on Weakly Supervised Learning @ ICLR 2021


May 8, 2021, Virtual





Deep learning relies on massive training sets of labeled examples to learn
from - often tens of thousands to millions to reach peak predictive
performance. However, large amounts of training data are only available for
very few standardized learning problems. Even small variations of the
problem specification or changes in the data distribution would necessitate
re-annotation of large amounts of data.

However, domain knowledge can often be expressed by sets of prototypical
descriptions: For example, vision experts can exploit meta information for
image labeling, linguists can describe discourse phenomena by prototypical
realization patterns, social scientists can specify events of interest by
characteristic key phrases, and bio-medical researchers have databases of
known interactions between drugs or proteins that can be used for heuristic
labeling. These knowledge-based descriptions can be either used as
rule-based predictors or as labeling functions for providing partial data
annotations. The growing field of weak supervision provides methods for
refining and generalizing such heuristic-based annotations in interaction
with deep neural networks and large amounts of unannotated data. In this
workshop, we want to advance theory, methods and tools for allowing experts
to express prior coded knowledge for automatic data annotations that can be
used to train arbitrary deep neural networks for prediction.




Topics of interest include, but are not limited to:

- Weak supervision in combination with neural networks and representation

- Theoretic insights into weak supervision

- Relationship between weak supervision and other machine learning
paradigms incl. semi-supervised learning, active learning and label

- Distant supervision and weak supervision for specific tasks

- Interdisciplinary applications of weak supervision

- Unification of weak supervision approaches from different fields, e.g,
relation extraction (natural language processing) and image classification

- Analysis of failure cases of weak supervision

- Benchmarks for evaluating and comparing weak supervision approaches

- Applications of weak supervision in industry settings




Feb 26, 2021: Paper Submission Deadline

Mar 26, 2021: Author Notification

May 8, 2021: Workshop Date

All deadlines are calculated at 11:59pm Pacific Daylight Savings Time (UTC
- 12h, Everywhere on Earth)




- Dan Roth <https://www.cis.upenn.edu/~danroth/>, University of Pennsylvania

- Paroma Varma <https://www.paroma.xyz/>, Snorkel AI

- Heng Ji <http://blender.cs.illinois.edu/hengji.html>, University of

- Lu Jiang <http://www.lujiang.info/>, Google Research

- Marine Carpuat <http://www.cs.umd.edu/~marine/>, University of Maryland




We solicit two categories of papers: long and short papers. Authors can
decide the archival status of their publications. Submissions will go
through a double-blind review process, where each submission is reviewed by
at least two program committee members.

Accepted papers will be presented by the authors either as talk or poster.
All submissions must follow the ICLR 2021 formatting requirements (

- Short paper submission: up to 4 pages of content (+1 on acceptance), plus

- Long paper submission: up to 8 pages of content (+1 on acceptance), plus




The co-chairs of the workshop can be contacted by email at:

weaksupervision (at) googlegroups.com




- Benjamin Roth, University of Vienna

- Barbara Plank, IT University of Copenhagen

- Alex Ratner, University of Washington

- Katharina Kann, University of Colorado Boulder

- Dietrich Klakow, Saarland University

- Michael A. Hedderich, Saarland University
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://visionscience.com/pipermail/visionlist_visionscience.com/attachments/20210107/5b940bc7/attachment-0001.html>

More information about the visionlist mailing list