Benchmarking LF-MMI, CTC and RNN-T Criteria for Streaming ASR
This provides a comprehensive benchmark for streaming ASR applications, particularly for social media video transcription across multiple languages, though it is incremental as it compares existing methods.
The authors benchmarked three training criteria (LF-MMI, CTC, RNN-T) for streaming ASR across 7 languages with 3K-14K hours of data, finding that RNN-T achieved the best accuracy while CTC was most efficient in inference.
In this work, to measure the accuracy and efficiency for a latency-controlled streaming automatic speech recognition (ASR) application, we perform comprehensive evaluations on three popular training criteria: LF-MMI, CTC and RNN-T. In transcribing social media videos of 7 languages with training data 3K-14K hours, we conduct large-scale controlled experimentation across each criterion using identical datasets and encoder model architecture. We find that RNN-T has consistent wins in ASR accuracy, while CTC models excel at inference efficiency. Moreover, we selectively examine various modeling strategies for different training criteria, including modeling units, encoder architectures, pre-training, etc. Given such large-scale real-world streaming ASR application, to our best knowledge, we present the first comprehensive benchmark on these three widely used training criteria across a great many languages.