Skip to main content

STOP: Spoken Task Oriented Parsing

A multi-domain compositional spoken language understanding dataset

Get Started on the Challenge

Updates

  • We have extended the deadline for the challenge to Feburary 13th 2023 and will be hosting submissions on Open Review, check out the submissions page
  • We just announced the call for participation for the Spoken Language Understanding (SLU) challenge @ ICASSP 2023, check out the call for participation
  • Our dataset paper was accepted at SLT 2022!
  • Experiment code is publicly available in fairseq!

Spoken Langauge Understanding Grand Challenge @ ICASSP 2023

We are excited to announce the Spoken Language Understanding (SLU) Grand Challenge @ ICASSP 2023. Participants of this challenge will compete in spoken language understanding across 3 tracks (1) quality (2) on-device and (3) low-resource and domain adaptation. To get started please head over to our page on ICASSP 2023 Grand Challenge!

Dataset Paper

To learn more details about SLU and Spoken Task Oriented Parsing (STOP) dataset. We encourage you to read our paper STOP: A dataset for Spoken Task Oriented Semantic Parsing. In order to get started with experimentation: all experiment code, baseline models, and pre-trained checkpoints are available in fairseq.