mlsanityData to 14 Oct 2026

Citrinet: Closing the Gap between Non-Autoregressive and Autoregressive End-to-End Models for Automatic Speech Recognition

arXiv 2104.01721 ↗First posted 5 Apr 2021eess.AScited by 34

Not on the map. Why?

Abstract (arXiv)

We propose Citrinet - a new end-to-end convolutional Connectionist Temporal Classification (CTC) based automatic speech recognition (ASR) model. Citrinet is deep residual neural model which uses 1D time-channel separable convolutions combined with sub-word encoding and squeeze-and-excitation. The resulting architecture significantly reduces the gap between non-autoregressive and sequence-to-sequence and transducer models. We evaluate Citrinet on LibriSpeech, TED-LIUM2, AISHELL-1 and Multilingual LibriSpeech (MLS) English speech datasets. Citrinet accuracy on these datasets is close to the best autoregressive Transducer models.

Key facts. Cited by 34 papers in the map (data through 14 Oct 2026). First posted on arXiv 5 Apr 2021.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 8; 2022: 10; 2023: 7; 2024: 4; 2025: 4; 2026: 1.

34 papers.

Sort:

By area:

  1. 2021SpeechBrain: A General-Purpose Speech Toolkitcited by 406
  2. 2022Squeezeformer: An Efficient Transformer for Automatic Speech Recognitioncited by 51
  3. 2021Efficient conformer: Progressive downsampling and grouped attention for automatic speech recognitioncited by 42
  4. 2021A Comparative Study on Non-Autoregressive Modelings for Speech-to-Text Generationcited by 28
  5. 2023Audio-Visual Efficient Conformer for Robust Speech Recognitioncited by 22
  6. 2021CTC Variations Through New WFST Topologiescited by 17
  7. 2021Hierarchical Conditional End-to-End ASR with CTC and Multi-Granular Subword Unitscited by 13
  8. 2022SoftCorrect: Error Correction with Soft Detection for Automatic Speech Recognitioncited by 9
  9. 2024VoiceShop: A Unified Speech-to-Speech Framework for Identity-Preserving Zero-Shot Voice Editingcited by 7
  10. 2021Decoupling recognition and transcription in Mandarin ASRcited by 6
  11. 2021A Toolbox for Construction and Analysis of Speech Datasetscited by 5
  12. 2022A Compact End-to-End Model with Local and Global Context for Spoken Language Identificationcited by 5
22 more
  1. 2022Inter-KD: Intermediate Knowledge Distillation for CTC-Based Automatic Speech Recognitioncited by 5
  2. 2022Improving CTC-based ASR Models with Gated Interlayer Collaborationcited by 4
  3. 2021Are E2E ASR models ready for an industrial usage?cited by 3
  4. 2022Seq-2-Seq based Refinement of ASR Output for Spoken Name Capturecited by 2
  5. 2023Dynamic Alignment Mask CTC: Improved Mask-CTC with Aligned Cross Entropycited by 2
  6. 2023Trustera: A Live Conversation Redaction Systemcited by 2
  7. 2023Enhancing Unsupervised Speech Recognition with Diffusion GANscited by 2
  8. 2024Non-autoregressive real-time Accent Conversion model with voice cloningcited by 2
  9. 2026A Comparative Study of Pretrained Transformer Models for Quranic ASR: Speech Representations, Label Formats, and Dataset Compositioncited by 2
  10. 2022ASR in German: A Detailed Error Analysiscited by 1
  11. 2023InterFormer: Interactive Local and Global Features Fusion for Automatic Speech Recognitioncited by 1
  12. 2023Weak Alignment Supervision from Hybrid Model Improves End-to-end ASRcited by 1
  13. 2025Visual-Aware Speech Recognition for Noisy Scenarioscited by 1
  14. 2022Improving Self-Supervised Learning-based MOS Prediction Networkscited by 0
  15. 2022Deep Sparse Conformer for Speech Recognitioncited by 0
  16. 2022Attention Enhanced Citrinet for Speech Recognitioncited by 0
  17. 2023E2E Spoken Entity Extraction for Virtual Agentscited by 0
  18. 2024Alternating Weak Triphone/BPE Alignment Supervision from Hybrid Model Improves End-to-End ASRcited by 0
  19. 2024PREBA: A Hardware/Software Co-Design for Multi-Instance GPU based AI Inference Serverscited by 0
  20. 2025RestoreGrad: Signal Restoration Using Conditional Denoising Diffusion Models with Jointly Learned Priorcited by 0
  21. 2025Remote Rowhammer Attack using Adversarial Observations on Federated Learning Clientscited by 0
  22. 2025E2E-VGuard: Adversarial Prevention for Production LLM-based End-To-End Speech Synthesiscited by 0