将Whisper模型压缩适配到树莓派，实现儿童语音的本地高效识别。
Adapting Whisper for Lightweight and Efficient Automatic Speech Recognition of Children for On-device Edge Applications
- 用低秩压缩和数据筛选优化tiny.en模型，适配边缘设备
- 在树莓派上实现15.9% WER，推理延迟仅0.23-0.41 RTF
- 模型体积减少51万参数，计算量降低2 GFLOPS，适合隐私敏感场景
由于监管与隐私问题，依赖云端进行儿童语音识别已面临挑战。本文旨在开发可在树莓派上运行的轻量高效Whisper ASR系统。基于MyST语料库，通过多种过滤策略微调`tiny.en'模型，实现15.9%的词错误率（过滤后为11.8%）。采用低秩压缩使编码器规模减少0.51M参数，在GPU上推理速度提升1.26倍，词错误率相对增加11%。在树莓派上推理时，压缩版本计算量减少约2 GFLOPS，两种模型的实时因子（RTF）均介于[0.23–0.41]之间。分析显示树莓派可支持小型模型运行，但小模型会引发额外开销与温度降频。
原文摘要 · Abstract (English)
Reliability on cloud providers for ASR inference to support child-centered voice-based applications is becoming challenging due to regulatory and privacy challenges. Motivated by a privacy-preserving design, this study aims to develop a lightweight & efficient Whisper ASR system capable of running on a Raspberry Pi. Upon evaluation of the MyST corpus and by examining various filtering strategies to fine-tune the `tiny.en' model, a Word Error Rate (WER) of 15.9% was achieved (11.8% filtered). A low-rank compression reduces the encoder size by 0.51M with 1.26x faster inference in GPU, with 11% relative WER increase. During inference on Pi, the compressed version required ~2 GFLOPS fewer computations. The RTF for both the models ranged between [0.23-0.41] for various input audio durations. Analyzing the RAM usage and CPU temperature showed that the PI was capable of handling both the tiny models, however it was noticed that small models initiated additional overhead/thermal throttling.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。