MSLG-SPA 2026 is the first shared task dedicated to bidirectional translation between Mexican Sign Language (MSL) glosses and Spanish, organized as part of IberLEF 2026. This task addresses a critical gap in the development of language technologies for Deaf communities in Mexico, where publicly available datasets, benchmarks, and computational resources for MSL remain extremely limited.
The task aims to establish a structured and reproducible evaluation framework for models that can both understand and generate MSL glosses, fostering the creation of future multimodal and inclusive translation pipelines. By bringing together researchers from natural language processing, sign language processing, and artificial intelligence, MSLG-SPA 2026 seeks to lay the foundation for a sustainable research ecosystem for Mexican Sign Language.
Participants are invited to develop and evaluate automatic translation systems (computational models) for:
Translating MSL gloss sequences into Spanish text.
Translating Spanish text into MSL gloss sequences.
This shared task promotes scientific progress, linguistic inclusion, and the long-term goal of accessible technologies for Deaf communities in Mexico and Spanish-speaking regions.