mlsanityData to 14 Oct 2026

SSVD: Structured SVD for Parameter-Efficient Fine-Tuning and Benchmarking under Domain Shift in ASR

arXiv 2509.02830 ↗First posted 2 Sep 2025cs.CL eess.AScited by 4

Not on the map. Why?

Abstract (arXiv)

Parameter-efficient fine-tuning (PEFT) has emerged as a scalable solution for adapting large foundation models. While low-rank adaptation (LoRA) is widely used in speech applications, its state-of-the-art variants, e.g., VeRA, DoRA, PiSSA, and SVFT, are developed mainly for language and vision tasks, with limited validation in speech. This work presents the first comprehensive integration and benchmarking of these PEFT methods within ESPnet. We further introduce structured SVD-guided (SSVD) fine-tuning, which selectively rotates input-associated right singular vectors while keeping output-associated vectors fixed to preserve semantic mappings. This design enables robust domain adaptation with minimal trainable parameters and improved efficiency. We evaluate all methods on domain-shifted speech recognition tasks, including child speech and dialectal variation, across model scales from 0.1B to 2B. All implementations are released in ESPnet to support reproducibility and future work.

Key facts. Cited by 4 papers in the map (data through 14 Oct 2026). Builds on 25 of its 33 references that are in the map. First posted on arXiv 2 Sep 2025.

Builds on

25 of 33 references.

References · 25

  1. 2017Attention Is All You Needcited by 60,361
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  3. 2019Parameter-Efficient Transfer Learning for NLPcited by 4,236
  4. 2022Robust Speech Recognition via Large-Scale Weak Supervisioncited by 4,057
  5. 2021The Power of Scale for Parameter-Efficient Prompt Tuningcited by 3,795
  6. 2021BitFit: Simple Parameter-efficient Fine-tuning for Transformer-based Masked Language-modelscited by 1,245
  7. 2024DoRA: Weight-Decomposed Low-Rank Adaptationcited by 758
  8. 2018ESPnet: End-to-End Speech Processing Toolkitcited by 558
  9. 2023VeRA: Vector-based Random Matrix Adaptationcited by 300
  10. 2024PiSSA: Principal Singular Values and Singular Vectors Adaptation of Large Language Modelscited by 269
  11. 2023QA-LoRA: Quantization-Aware Low-Rank Adaptation of Large Language Modelscited by 155
  12. 2021Orthogonalizing Convolutional Layers with the Cayley Transformcited by 105
13 more
  1. 2022E-Branchformer: Branchformer with Enhanced merging for speech recognitioncited by 92
  2. 2024OWSM v3.1: Better and Faster Open Whisper-Style Speech Models based on E-Branchformercited by 74
  3. 2024SVFT: Parameter-Efficient Fine-Tuning with Singular Vectorscited by 58
  4. 2023Reproducing Whisper-Style Training Using an Open-Source Toolkit and Publicly Available Datacited by 42
  5. 2024LoRA-Whisper: Parameter-Efficient and Extensible Multilingual ASRcited by 36
  6. 2024Benchmarking Children's ASR with Supervised and Self-supervised Speech Foundation Modelscited by 29
  7. 2023My Science Tutor (MyST) -- A Large Corpus of Children's Conversational Speechcited by 28
  8. 2023Kid-Whisper: Towards Bridging the Performance Gap in Automatic Speech Recognition for Children VS. Adultscited by 27
  9. 2023Sparsely Shared LoRA on Whisper for Child Speech Recognitioncited by 18
  10. 2025OWLS: Scaling Laws for Multilingual Speech Recognition and Translation Modelscited by 17
  11. 2023Leveraging Large Language Models for Exploiting ASR Uncertaintycited by 15
  12. 2024Towards Rehearsal-Free Multilingual ASR: A LoRA-based Case Study on Whispercited by 9
  13. 2022Bottleneck Low-rank Transformers for Low-resource Spoken Language Understandingcited by 6

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 4.

4 papers.

  1. 2026A Language-Agnostic Hierarchical LoRA-MoE Architecture for CTC-based Multilingual ASRcited by 1
  2. 2026SSVD-O: Parameter-Efficient Fine-Tuning with Structured SVD for Speech Recognitioncited by 1
  3. 2026AntiPaSTO: Self-Supervised Honesty Steering via Anti-Parallel Representationscited by 0
  4. 2026Frozen Factor or Spectral Band? Disentangling Two Choices in Low-Rank LoRAcited by 0