Speechnet: Weakly Supervised, End-to-end Speech Recognition At Industrial Scale

Abstract

End-to-end automatic speech recognition systems represent the state of the art, but they rely on thousands of hours of manually annotated speech for training, as well as heavyweight computation for inference. Of course, this impedes commercialization since most companies lack vast human and computational resources. In this paper, we explore training and deploying an ASR system in the label-scarce, compute-limited setting. To reduce human labor, we use a third-party ASR system as a weak supervision source, supplemented with labeling functions derived from implicit user feedback. To accelerate inference, we propose to route production-time queries across a pool of CUDA graphs of varying input lengths, the distribution of which best matches the traffic's. Compared to our third-party ASR, we achieve a relative improvement in word-error rate of 8% and a speedup of 600%. Our system, called SpeechNet, currently serves 12 million queries per day on our voice-enabled smart television. To our kn

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keytang2022speechnet

Related papers

• Fast Offline Transformer-based End-to-end Automatic Speech Recognition For Real-world Applications (2021)7.16
• Unified End-to-end Speech Recognition And Endpointing For Fast And Efficient Speech Systems (2022)5.24
• From Weak Labels To Strong Results: Utilizing 5,000 Hours Of Noisy Classroom Transcripts With Minimal Accurate Data (2025)0.00
• Towards A Competitive End-to-end Speech Recognition For Chime-6 Dinner Party Transcription (2020)6.77
• Bigssl: Exploring The Frontier Of Large-scale Semi-supervised Learning For Automatic Speech Recognition (2021)15.73
• Advances In All-neural Speech Recognition (2016)11.29
• Label-synchronous Neural Transducer For Adaptable Online E2E Speech Recognition (2023)3.58
• Less Is More: Accurate Speech Recognition & Translation Without Web-scale Data (2024)0.00