We are pleased to announce the upcoming full-day SPEAKABLE Workshop on Speech Language Models in Low-Resource Settings: Performance, Evaluation, and Bias Analysis, co-located with LREC 2026. This workshop brings together researchers, practitioners, and industry experts working to advance speech technology for under-resourced languages. We invite contributions that address the unique challenges and opportunities in this space.

More information at https://speakable-2026.github.io/

Workshop Overview:

Speech Language Models have become foundational in speech and spoken language processing, yet their performance remains uneven across languages and communities. Models trained at scale often struggle in low-resource settings due to limited annotated data, domain mismatch, and sociolinguistic variation, while also raising concerns related to bias, robustness, and evaluation reliability. These challenges are particularly pronounced for under-represented languages, where errors and biases can have disproportionate downstream impact.

The goal of the SPEAKABLE workshop is to bring together researchers working on speech language technologies for low-resource settings, with a focus on performance, evaluation, and bias analysis. We welcome contributions that investigate, develop, or critically assess models, datasets, benchmarks, and methodologies aimed at improving speech technologies under data scarcity, enhancing fairness and inclusivity, and establishing meaningful evaluation practices for under-resourced languages.

Topics of Interest:
We encourage submissions on (but not limited to):
- Performance of speech language models in low-resource and underrepresented languages
- Evaluation methodologies and creation of benchmarks
- Bias analysis, detection, and mitigation strategies in speech technologies
- Real-world applications, deployment challenges, and case studies
- Speech recognition, speech-to-text, language modeling, multilingual and cross-lingual approaches
- Fairness, ethical considerations, and inclusive NLP for low-resource speech communities

We welcome original research, position papers, and ongoing work relevant to speech and language modeling for low-resource settings.

Submission Guidelines:
Submission format: Papers upto 4 (short papers) and 8 (long papers) pages excluding references.
Style: All submissions must follow the LREC 2026 format and use the official LREC author kit. (available at https://lrec2026.info/authors-kit/ ).
Review Process: Double-blind peer review. Submissions must be fully anonymized.
Submission system: Papers must be submitted via the following link: https://softconf.com/lrec2026/SPEAKABLE2026/
Language Resources: In line with LREC policies, authors are encouraged to describe, document, and share language resources, datasets, models, evaluation tools, or annotation guidelines used or created in their work.
Accepted papers: All accepted papers will be included in the LREC 2026 workshop proceedings.
Presentation: Accepted papers will be presented as oral or poster sessions during the workshop.

Important Dates:
All deadlines are 11:59PM UTC-12:00 (“anywhere on Earth”)
Paper Submission Deadline: 16 February 2026
Notification of Acceptance: 12 March 2026
Camera-Ready Papers: 30 March 2026
Workshop Date: 11 May 2026

Contact:
For questions, please contact the workshop organizers at: speakable2026@gmail.com

Organizing Committee:
Nina Hosseini-Kivanani (RTL & University of Luxembourg, Luxembourg)
Alessio Brutti (Fondazione Bruno Kessler, Italy)
Marco Matassoni (Fondazione Bruno Kessler, Italy)
Sandipana Dowerah (Tallinn University of Technology, Estonia)
Davide Liga (University of Luxembourg, Luxembourg)
Christoph Schommer (University of Luxembourg, Luxembourg)