Shared Task 2: Multilingual ASR for South Asian Languages
South Asia hosts one of the world’s richest concentrations of linguistic diversity, yet many of its languages remain underrepresented in speech technologies. Differences in resource availability, scripts, phonology, and domains pose significant challenges to building robust Automatic Speech Recognition (ASR) systems that generalize beyond high-resource settings. This shared task aims to advance multilingual and cross-lingual ASR for South Asian languages by encouraging the development of accurate, efficient, and generalizable models.
The task focuses on eight South Asian languages drawn from different language families and regions: Balti, Bengali, Dhivehi, Marathi, Nepali, Saraiki, Tamil, and Torwali. These languages use different scripts and represent a spectrum of high-, medium-, and low-resource conditions. The training, development, and evaluation data will be derived from the Mozilla Common Voice v23 dataset, ensuring a uniform and openly available data source.
Sub Task 1: Multilingual ASR
The core mandatory subtask is multilingual ASR, in which participants are required to build a single system capable of recognizing speech in all eight languages. To promote both performance and deployability, each team must submit two model variants:
a small model with fewer than 400 million parameters, and
a large model with more than 900 million parameters.
It explicitly encourages research on scalability, efficiency, and performance trade-offs in multilingual ASR. Systems will be evaluated using Word Error Rate (WER) and Character Error Rate (CER) on a per-language basis, with macro-averaged CER as the primary metric.
Sub Task 2: ASR for an Unseen Language
The optional subtask focuses on zero-shot and few-shot ASR, using the multilingual ASR system developed above, for an unseen South Asian language that is not included in the multilingual training set. The evaluation metric for this subtask will be CER.
By bringing together multilingual ASR, model efficiency, and cross-lingual generalization, this shared task aims to foster inclusive speech technologies and advance ASR research for underrepresented South Asian languages.
All deadlines are 11:59 PM UTC-12:00 (“anywhere on Earth”).
First CFP: 15 December 2025
Dataset for Sub Task 1 available: 15 December, 2025
Dataset for Sub Task 2 available: 03 February, 2025
Submission Deadline: 20 February 2026
Notification of Acceptance: 20 March 2026