Challenge: ASR models can be used to correct accent-specific errors without ground truth . pseudo-labels inherit the teacher model's systematic biases, authors say .
Approach: They propose a parameter-space correction technique that captures pseudo-label biases . they propose achieving up to 35% relative WER reduction on a pseudo-labeled target model .
Outcome: The proposed model achieves 35% relative WER reduction on ten African accents with the Whisper tiny model.

Similar Papers

Joint Speech Transcription and Translation: Pseudo-Labeling with Out-of-Distribution Data (2023.findings-acl)

Copied to clipboard

Challenge: a recent study shows that self-training can improve upon fully supervised baselines in low-resource settings for several sequence-to-sequence tasks.
Approach: They propose to use pseudo-labeling to label unsupervised data and add it to the training pool.
Outcome: The proposed setup improves on the unsupervised data by using pseudo-labeling . the proposed setup provides 0.4% absolute WER and 2.1 BLEU points for En–De .
Task Arithmetic can Mitigate Synthetic-to-Real Gap in Automatic Speech Recognition (2024.emnlp-main)

Copied to clipboard

Challenge: Existing methods for speech recognition suffer from the synthetic-to-real gap . existing methods suffer from this distributional shift due to acoustic mismatches .
Approach: They propose to use task arithmetic to fine-tune an ASR model on synthetic data to mitigate the synthetic-to-real gap.
Outcome: The proposed method shows an improvement of 10.03% over baselines on the SLURP dataset.
On the Robust Approximation of ASR Metrics (2025.findings-acl)

Copied to clipboard

Challenge: Existing methods for estimating speech recognition metrics depend on ground truth labels.
Approach: They propose a label-free approach to approximating ASR performance metrics . they embed multimodal embeddings in a unified space for speech and transcription representations .
Outcome: The proposed method outperforms baseline models on speech recognition benchmarks by 50%.
Why Aren’t We NER Yet? Artifacts of ASR Errors in Named Entity Recognition in Spontaneous Speech Transcripts (2023.acl-long)

Copied to clipboard

Challenge: despite advances in language models, the transcript of spontaneous human-human conversations remains an insurmountable challenge for most models.
Approach: They examine the relationship between ASR and NER errors which limit NER models' ability to recover entity mentions from spontaneous speech transcripts.
Outcome: The proposed model fails even if no word errors are introduced by the ASR . the proposed model's performance deteriorates when applied to the ASL outputs .
Looking Beyond Label Noise: Shifted Label Distribution Matters in Distantly Supervised Relation Extraction (D19-1)

Copied to clipboard

Challenge: Existing studies on DS-based relation extraction (RE) methods focus on handling label noise, but other factors may have been overlooked.
Approach: They propose a method to automatically adjust DS-RE models to a shifted label distribution problem . they find this problem exists in real-world DS datasets and can be overcome .
Outcome: The proposed method achieves consistent performance gains on DS-trained models with an up to 23% relative F1 improvement, which verifies their assumptions.
From Fake to Real: Mitigating Out-of-Distribution Bias in In-Context Learning via Feedback Supervision from Large Language Models (2026.findings-acl)

Copied to clipboard

Challenge: In-Context Learning (ICL) is one of the most common methods for complex Natural Language Understanding tasks.
Approach: They propose a method that uses model confidence and perturbation perplexity to enhance the quality of pseudo-labels.
Outcome: The proposed method reduces OOD biases by avoiding direct use of source data.
An Empirical Study of Incorporating Pseudo Data into Grammatical Error Correction (D19-1)

Copied to clipboard

Challenge: incorporating pseudo data in the training of grammatical error correction models has been a key factor in improving performance of such models.
Approach: They investigate the choice of how pseudo data should be generated or used in a grammatical error correction model and show that the results are state-of-the-art.
Outcome: The proposed method achieves state-of-the-art on the CoNLL-2014 test set and the official test set of the BEA-2019 shared task without making any modifications to the model architecture.
Evaluating Open-Source ASR Systems: Performance Across Diverse Audio Conditions and Error Correction Methods (2025.coling-main)

Copied to clipboard

Challenge: Automated speech recognition (ASR) systems are able to transcribe spontaneous human conversations with high accuracy.
Approach: They evaluate the accuracy of open source automatic speech recognition systems across conversational speech datasets and explore the potential of ASR ensembling and post-ASR correction methods to improve transcription accuracy.
Outcome: The proposed methods highlight the need for robust error correction techniques and address demographic biases to enhance ASR performance and inclusivity.
PCAD: Towards ASR-Robust Spoken Language Understanding via Prototype Calibration and Asymmetric Decoupling (2024.acl-long)

Copied to clipboard

Challenge: Spoken language understanding (SLU) suffers from error propagation from automatic speech recognition (ASR) in actual scenarios.
Approach: They propose a framework which calibrates bias and errors and achieves adaptive-balanced decoupling training by a prototype-based loss model.
Outcome: The proposed framework outperforms existing approaches and achieves state-of-the-art performance on three datasets.
NoiseBench: Benchmarking the Impact of Real Label Noise on Named Entity Recognition (2024.emnlp-main)

Copied to clipboard

Challenge: Existing approaches to named entity recognition often contain a significant percentage of incorrect labels for entity types and boundary boundaries.
Approach: They propose a noise-robust learning approach that learns from data with partially incorrect labels.
Outcome: The proposed methods are based on simulated noise and are easier to handle than simulated real noise caused by human error or semi-automatic annotation.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations