Moonshine: Speech Recognition for Live Transcription and Voice Commands
This work addresses the need for real-time, resource-constrained speech recognition applications, representing an incremental improvement over existing methods.
The paper tackles the problem of efficient speech recognition for live transcription and voice commands by introducing Moonshine, a transformer-based model that reduces compute requirements by 5x compared to Whisper tiny-en while maintaining word error rates.
This paper introduces Moonshine, a family of speech recognition models optimized for live transcription and voice command processing. Moonshine is based on an encoder-decoder transformer architecture and employs Rotary Position Embedding (RoPE) instead of traditional absolute position embeddings. The model is trained on speech segments of various lengths, but without using zero-padding, leading to greater efficiency for the encoder during inference time. When benchmarked against OpenAI's Whisper tiny-en, Moonshine Tiny demonstrates a 5x reduction in compute requirements for transcribing a 10-second speech segment while incurring no increase in word error rates across standard evaluation datasets. These results highlight Moonshine's potential for real-time and resource-constrained applications.