Renaissance Labs

Atlas / Europe

Academic lab · Espoo, Finland

Aalto Speech Recognition

Aalto University's speech recognition group, whose open Morfessor and AaltoASR tools have long supported speech and language technology for morphologically rich languages.

speechmorphology

Espoo, Finland

Type
Academic lab
Based in
Espoo, Finland
Featured projects
4

About

The speech recognition group sits in the Department of Information and Communications Engineering at Aalto University in Espoo and is part of the Finnish Center for Artificial Intelligence. It works on acoustic modeling of phonemes, statistical language models for large-vocabulary continuous speech recognition, and speech technology for under-resourced languages.

The group publishes its research software through the aalto-speech GitHub organization, created in 2013. The best known repositories are Morfessor, a tool for unsupervised and semi-supervised morphological segmentation, AaltoASR, a set of recognition tools, and speaker diarization scripts built on that toolkit.

Professor Mikko Kurimo heads the group and is known internationally for unsupervised subword modeling for languages with rich word inflection, such as Finnish, Estonian, Turkish and Arabic. The software mainly serves other speech and language researchers.

Mission and ambitions

The group aims to improve speech recognition with new algorithms and to make recognition practical for languages with limited data and heavy inflection. It shares its tools openly so that other researchers can reuse and extend them.

Known for

  1. 01

    Morfessor morphological segmentation

  2. 02

    AaltoASR recognition toolkit

  3. 03

    Subword modeling for inflected languages

  4. 04

    Open speaker diarization scripts

Milestones

  1. 2013

    Morfessor repository

    The Morfessor tool for unsupervised and semi-supervised morphological segmentation appeared on the group's GitHub organization under a BSD licence.

  2. 2013

    AaltoASR toolkit

    The group published the Aalto Automatic Speech Recognition tools as an open repository.

  3. 2016

    Speaker diarization scripts

    Diarization scripts based on AaltoASR were released and have drawn close to 200 GitHub stars.

  4. 2018

    Subword Kaldi recipes

    A repository showing how to handle position-dependent phones in a subword lexicon FST for Kaldi was published.

Open source

GitHubgithub.com/aalto-speech
  • Morfessor is a tool for unsupervised and semi-supervised morphological segmentation (207 stars)

  • Speaker diarization scripts, based on AaltoASR (192 stars)

  • Aalto Automatic Speech Recognition tools (90 stars)

  • Properly handle position-dependent phones in a subword lexicon FST (31 stars)

Notable releases

Openness

Tools are open source on GitHub.

People

Listed from the lab's own pages and public profiles.

Contact and community

Contact via site.

Sources

Compiled from the lab's public pages, Hugging Face and GitHub. Last checked 2026-11-17. Something wrong or out of date? Tell us on Discord.

Labs nearby