mlsanityData to 7 Oct 2026

NIRANTAR: Continual Learning with New Languages and Domains on Real-world Speech Data

arXiv 2507.00534 ↗First posted 1 Jul 2025cs.CLcited by 2

Not on the map. Why?

Abstract (arXiv)

We introduce Nirantar, a comprehensive framework for evaluating continual learning (CL) in multilingual and multi-domain ASR. Designed to reflect real-world CL challenges, Nirantar leverages data collected incrementally across 22 languages and 208 districts in India through natural episodes. This enables evaluation across Language-Incremental (LIL), Domain-Incremental (DIL), and the novel Language-Incremental Domain-Incremental Learning (LIDIL) scenarios. Unlike prior work that relies on simulated episodes, Nirantar presents dynamic, non-uniform language and domain shifts, making it an ideal testbed for CL research. With 3250 hours of human-transcribed speech, including 1720 hours newly introduced in this work, our framework enables systematic benchmarking of CL methods. We evaluate existing approaches and demonstrate that no single method performs consistently well, underscoring the need for more robust CL strategies.

Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 37 of its 48 references that are in the map. First posted on arXiv 1 Jul 2025.

Builds on

37 of 48 references.

References · 37

  1. 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 3,972
  2. 2020Conformer: Convolution-augmented Transformer for Speech Recognitioncited by 1,838
  3. 2019Common Voice: A Massively-Multilingual Speech Corpuscited by 1,358
  4. 2016Progressive Neural Networkscited by 1,332
  5. 2017Memory Aware Synapses: Learning what (not) to forgetcited by 1,094
  6. 2018Experience Replay for Continual Learningcited by 1,074
  7. 2018Efficient Lifelong Learning with A-GEMcited by 1,044
  8. 2020Dark Experience for General Continual Learning: a Strong, Simple Baselinecited by 895
  9. 2017PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruningcited by 882
  10. 2023A Comprehensive Survey of Continual Learning: Theory, Method and Applicationcited by 860
  11. 2013An Empirical Investigation of Catastrophic Forgetting in Gradient-Based Neural Networkscited by 711
  12. 2017Learning multiple visual domains with residual adapterscited by 553
25 more
  1. 2021VoxPopuli: A Large-Scale Multilingual Speech Corpus for Representation Learning, Semi-Supervised Learning and Interpretationcited by 455
  2. 2017CORe50: a New Dataset and Benchmark for Continuous Object Recognitioncited by 287
  3. 2023Google USM: Scaling Automatic Speech Recognition Beyond 100 Languagescited by 228
  4. 2020Overcoming Catastrophic Forgetting in Graph Neural Networkscited by 93
  5. 2022The CLEAR Benchmark: Continual LEArning on Real-World Imagerycited by 92
  6. 2020A Wholistic View of Continual Learning with Deep Neural Networks: Forgotten Lessons and the Bridge to Active and Open World Learningcited by 86
  7. 2021Multilingual and code-switching ASR challenges for low resource Indian languagescited by 50
  8. 2024IndicVoices: Towards building an Inclusive Multilingual Speech Dataset for Indian Languagescited by 39
  9. 2022IndicSUPERB: A Speech Processing Universal Performance Benchmark for Indian languagescited by 38
  10. 2021Learn Continually, Generalize Rapidly: Lifelong Knowledge Accumulation for Few-shot Learningcited by 35
  11. 2021Pseudo-Labeling for Massively Multilingual Speech Recognitioncited by 23
  12. 2023Vistaar: Diverse Benchmarks and Training Sets for Indian Language ASRcited by 21
  13. 2021Towards Lifelong Learning of End-to-end ASRcited by 20
  14. 2023Stateful Conformer with Cache-based Inference for Streaming Automatic Speech Recognitioncited by 20
  15. 2021Uncertainty-Aware Balancing for Multilingual and Multi-Domain Neural Machine Translation Trainingcited by 19
  16. 2022Effectiveness of Mining Audio and Text Pairs from Public Data for Improving ASR Systems for Low-Resource Languagescited by 16
  17. 2023CL-MASR: A Continual Learning Benchmark for Multilingual ASRcited by 16
  18. 2022Using Adapters to Overcome Catastrophic Forgetting in End-to-End Automatic Speech Recognitioncited by 15
  19. 2023Svarah: Evaluating English ASR Systems on Indian Accentscited by 15
  20. 2021Automatic Speech Recognition in Sanskrit: A New Speech Corpus and Modelling Insightscited by 14
  21. 2022Subword Dictionary Learning and Segmentation Techniques for Automatic Speech Recognition in Tamil and Kannadacited by 10
  22. 2023SPRING-INX: A Multilingual Indian Language Speech Corpus by SPRING Lab, IIT Madrascited by 9
  23. 2024LAHAJA: A Robust Multi-accent Benchmark for Evaluating Hindi ASR Systemscited by 8
  24. 2024Multi-Label Continual Learning for the Medical Domain: A Novel Benchmarkcited by 6
  25. 2023SPIRE-SIES: A Spontaneous Indian English Speech Corpuscited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 1; 2026: 1.

2 papers.

  1. 2025Navigating the Reality Gap: On-Device Continual Adaptation of ASR for Clinical Telephonycited by 0
  2. 2026Adapting Text LLMs to Speech via Multimodal Depth Up-Scalingcited by 0