mlsanityData to 19 Oct 2026

CLAR: CIF-Localized Alignment for Retrieval-Augmented Speech LLM-Based Contextual ASR

arXiv 2603.25460 ↗First posted 26 Mar 2026cs.SDcited by 1

On the map: Speech and audio language models → Speech recognition models and training

Abstract (arXiv)

Speech LLM-based ASR often struggles with named entities and long-tail words due to strong internal language-model priors. Retrieval-augmented biasing can help, but its effectiveness depends on accurate hotword localization in full-utterance speech under weak supervision. We propose CLAR, a dual-encoder speech-text retriever that uses Continuous Integrate-and-Fire (CIF) to learn monotonic token-level alignments without timestamps. With length-aware localized matching, CLAR anchors short-entity acoustic cues and reduces representation dilution and attention drift. The retriever is trained with a multi-granularity objective combining global and local segment-level contrastive losses and a CIF quantity constraint. At inference, top-ranked hotwords are injected as contextual prompts for the Speech LLM, improving recognition without shallow fusion. Experiments show that CLAR significantly improves hotword retrieval and reduces both CER and B-WER against strong contextual ASR baselines.

Key facts. Cited by 1 paper in the map (data through 19 Oct 2026). Placed in Speech recognition models and training, one of 11 directions in Speech and audio language models. Builds on 25 of its 28 references that are in the map. First posted on arXiv 26 Mar 2026.

Builds on

25 of 28 references.

Closest ancestors · 17

Same direction, newest first

  1. 2026Retrieval-Augmented Self-Taught Reasoning Model with Adaptive Chain-of-Thought for ASR Named Entity Correction21 Jan
  2. 2025Retrieval Augmented Generation based context discovery for ASR23 Sep
  3. 2025PAC: Pronunciation-Aware Contextualized Large Language Model-based Automatic Speech Recognition16 Sep
  4. 2025Fun-ASR Technical Report15 Sep
  5. 2025Generative Annotation for ASR Named Entity Correction28 Aug
  6. 2025Improving Speech Recognition of Named Entities in Classroom Speech with LLM Revision and Phonetic-Semantic Context12 Jun
  7. 2025BR-ASR: Efficient and Scalable Bias Retrieval Framework for Contextual Biasing ASR in Speech LLM25 May
  8. 2025FireRedASR: Open-Source Industrial-Grade Mandarin Speech Recognition Models from Encoder-Decoder to LLM Integration24 Jan
  9. 2024CTC-Assisted LLM-Based Contextual ASR10 Nov
  10. 2024Efficient Streaming LLM for Speech Recognition2 Oct
  11. 2024Contextualization of ASR with LLM using phonetic retrieval-based augmentation11 Sep
  12. 2024Retrieval Augmented Correction of Named Entity Speech Recognition Errors9 Sep
5 more
  1. 2024Seed-ASR: Understanding Diverse Speech and Contexts with LLM-based Speech Recognition5 Jul
  2. 2023Contextual Biasing of Named-Entities with Large Language Models1 Sep
  3. 2023SeACo-Paraformer: A Non-Autoregressive ASR System with Flexible and Effective Hotword Customization Ability7 Aug
  4. 2022Paraformer: Fast and Accurate Parallel Transformer for Non-autoregressive End-to-End Speech Recognition16 Jun
  5. 2021Contextualized Streaming End-to-End Speech Recognition with Trie-Based Deep Biasing and Shallow Fusion5 Apr

Other references · 8

  1. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,833
  2. 2017AISHELL-1: An Open-Source Mandarin Speech Corpus and A Speech Recognition Baselinecited by 497
  3. 2018AISHELL-2: Transforming Mandarin ASR Research Into Industrial Scalecited by 181
  4. 2026Qwen3-ASR Technical Reportcited by 162
  5. 2025Step-Audio 2 Technical Reportcited by 142
  6. 2019CIF: Continuous Integrate-and-Fire for End-to-End Speech Recognitioncited by 104
  7. 2023Natural Language Supervision for General-Purpose Audio Representationscited by 75
  8. 2025Leveraging LLM for Stuttering Speech: A Unified Architecture Bridging Recognition and Event Detectioncited by 3

Built on by

Cited by year (2026 to 19 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Training-Free Contextual ASR via SpeechLLM-Based Error-Aware Selective Retrievalcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Contextual Biasing for LLM-Based ASR with Hotword Retrieval and Reinforcement Learning
  2. 2025M-CIF: Multi-Scale Alignment For CIF-Based Non-Autoregressive ASR
  3. 2025CMT-LLM: Contextual Multi-Talker ASR Utilizing Large Language Models
  4. 2025End-to-end Contrastive Language-Speech Pretraining Model For Long-form Spoken Question Answering
  5. 2024A Comprehensive Solution to Connect Speech Encoder and Large Language Model for ASR
  6. 2026Gradient-Based Speech-to-Text Alignment for Any ASR Model: From CTC to Speech LLMs
  7. 2026SSEmb: Training a Generalizable Speech Retrieval Model Using Only Automatic Speech Recognition Data
  8. 2025SEAL: Speech Embedding Alignment Learning for Speech Large Language Model with Retrieval-Augmented Generation