REVIEW 5 cited by
Anatomy of Industrial Scale Multilingual ASR
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
This paper describes AssemblyAI's industrial-scale automatic speech recognition (ASR) system, designed to meet the requirements of large-scale, multilingual ASR serving various application needs. Our system leverages a diverse training dataset comprising unsupervised (12.5M hours), supervised (188k hours), and pseudo-labeled (1.6M hours) data across four languages. We provide a detailed description of our model architecture, consisting of a full-context 600M-parameter Conformer encoder pre-trained with BEST-RQ and an RNN-T decoder fine-tuned jointly with the encoder. Our extensive evaluation demonstrates competitive word error rates (WERs) against larger and more computationally expensive models, such as Whisper large and Canary-1B. Furthermore, our architectural choices yield several key advantages, including an improved code-switching capability, a 5x inference speedup compared to an optimized Whisper baseline, a 30% reduction in hallucination rate on speech data, and a 90% reduction in ambient noise compared to Whisper, along with significantly improved time-stamp accuracy. Throughout this work, we adopt a system-centric approach to analyzing various aspects of fully-fledged ASR models to gain practically relevant insights useful for real-world services operating at scale.
Forward citations
Cited by 5 Pith papers
-
Early Attentive Sparsification Accelerates Neural Speech Transcription
Attention-based early audio-token sparsification at 40-60% sparsity accelerates Whisper ASR up to 1.6x with under 1% relative WER loss, across ten model variants, with no fine-tuning.
-
SC-SOT: Conditioning the Decoder on Diarized Speaker Information for End-to-End Overlapped Speech Recognition
Conditioning an SOT multi-talker ASR decoder on EEND-EDA speaker embeddings and activity information lowers WER on Libri2Mix and Libri3Mix, provided the diarization branch is accurate.
-
VietASR: Achieving Industry-level Vietnamese ASR with 50-hour labeled data and Large-Scale Speech Pretraining
A 68M-parameter Vietnamese ASR model, pretrained on 70,000 hours of unlabeled audio and fine-tuned on 50 hours of labels, reports average WER 8.31, beating Whisper Large-v3 and commercial systems.
-
Selective Invocation for Multilingual ASR: A Cost-effective Approach Adapting to Speech Recognition Difficulty
A spoken large language model learns to judge its own transcription difficulty and routes only hard speech to a stronger ASR model, cutting cost and improving word error rate.
-
SHNU Multilingual Conversational Speech Recognition System for INTERSPEECH 2025 MLC-SLM Challenge
SHNU-mASR, a parallel-encoder LLM system, achieves 11.76% CER/WER on the MLC-SLM blind eval set, 8.41 points better than the official baseline.
Discussion (0). Continue with ORCID to comment.