{
  "source_file": "0930",
  "source_platform": "研究机构",
  "author": "OpenAI Blog",
  "title_listed": "Whisper: Robust Speech Recognition via Large-Scale Weak Supervision",
  "keywords": "Whisper,语音识别,弱监督",
  "domain": "文本处理",
  "content_type": "深度技术解析",
  "is_paid_flag": "否",
  "url": "https://openai.com/research/whisper",
  "publish_date": "2022-09",
  "corpus_stage": "预训练/SFT",
  "notes": "Whisper语音识别模型",
  "id": "f5b4364f5d61",
  "collected_at": "2026-10-02T15:28:53",
  "title": "Introducing Whisper",
  "body_markdown": "Introducing Whisper\nWe’ve trained and are open-sourcing a neural net called Whisper that approaches human level robustness and accuracy on English speech recognition.\nThis post introduced Whisper in 2022.\nFor current audio and transcription capabilities:\nWhisper is an automatic speech recognition (ASR) system trained on 680,000 hours of multilingual and multitask supervised data collected from the web. We show that the use of such a large and diverse dataset leads to improved robustness to accents, background noise and technical language. Moreover, it enables transcription in multiple languages, as well as translation from those languages into English. We are open-sourcing models and inference code to serve as a foundation for building useful applications and for further research on robust speech processing.\nThe Whisper architecture is a simple end-to-end approach, implemented as an encoder-decoder Transformer. Input audio is split into 30-second chunks, converted into a log-Mel spectrogram, and then passed into an encoder. A decoder is trained to predict the corresponding text caption, intermixed with special tokens that direct the single model to perform tasks such as language identification, phrase-level timestamps, multilingual speech transcription, and to-English speech translation.\nOther existing approaches frequently use smaller, more closely paired audio-text training datasets, , or use broad but unsupervised audio pretraining., , Because Whisper was trained on a large and diverse dataset and was not fine-tuned to any specific one, it does not beat models that specialize in LibriSpeech performance, a famously competitive benchmark in speech recognition. However, when we measure Whisper’s zero-shot performance across many diverse datasets we find it is much more robust and makes 50% fewer errors than those models.\nAbout a third of Whisper’s audio dataset is non-English, and it is alternately given the task of transcribing in the original language or translating to English. We find this approach is particularly effective at learning speech to text translation and outperforms the supervised SOTA on CoVoST2 to English translation zero-shot.\nWe hope Whisper’s high accuracy and ease of use will allow developers to add voice interfaces to a much wider set of applications. Check out the [paper(opens in a new window)](https://cdn.openai.com/papers/whisper.pdf), [model card(opens in a new window)](https://github.com/openai/whisper/blob/main/model-card.md), and [code(opens in a new window)](https://github.com/openai/whisper) to learn more details and to try out Whisper.\nReferences\nRelated articles\n[View all](/news/)",
  "metadata": {
    "title": "Introducing Whisper",
    "engine": "curl_cffi"
  },
  "paywall": {
    "listed_paid": false,
    "detected": false,
    "marker": "",
    "public_only": false,
    "policy": "付费/会员内容只保留页面公开可见部分，不绕过付费墙"
  },
  "crawl": {
    "success": true,
    "status_code": 200,
    "error": "",
    "retried": true,
    "engine": "curl_cffi+trafilatura"
  },
  "code_blocks": [],
  "images": [],
  "audio": [],
  "videos": [],
  "links": [],
  "retry_at": "2026-10-03T08:11:51",
  "final_url": "https://openai.com/index/whisper/",
  "repair_version": 2,
  "repaired_at": "2026-10-20T17:31:02Z",
  "source_url": "https://openai.com/research/whisper",
  "delivery_status": "unverified",
  "discovery_status": "http_403",
  "missing_reason": "http_403",
  "repair_error": "",
  "actual_video_files": [],
  "source_body_markdown": "Introducing Whisper\nWe’ve trained and are open-sourcing a neural net called Whisper that approaches human level robustness and accuracy on English speech recognition.\nThis post introduced Whisper in 2022.\nFor current audio and transcription capabilities:\nWhisper is an automatic speech recognition (ASR) system trained on 680,000 hours of multilingual and multitask supervised data collected from the web. We show that the use of such a large and diverse dataset leads to improved robustness to accents, background noise and technical language. Moreover, it enables transcription in multiple languages, as well as translation from those languages into English. We are open-sourcing models and inference code to serve as a foundation for building useful applications and for further research on robust speech processing.\nThe Whisper architecture is a simple end-to-end approach, implemented as an encoder-decoder Transformer. Input audio is split into 30-second chunks, converted into a log-Mel spectrogram, and then passed into an encoder. A decoder is trained to predict the corresponding text caption, intermixed with special tokens that direct the single model to perform tasks such as language identification, phrase-level timestamps, multilingual speech transcription, and to-English speech translation.\nOther existing approaches frequently use smaller, more closely paired audio-text training datasets, , or use broad but unsupervised audio pretraining., , Because Whisper was trained on a large and diverse dataset and was not fine-tuned to any specific one, it does not beat models that specialize in LibriSpeech performance, a famously competitive benchmark in speech recognition. However, when we measure Whisper’s zero-shot performance across many diverse datasets we find it is much more robust and makes 50% fewer errors than those models.\nAbout a third of Whisper’s audio dataset is non-English, and it is alternately given the task of transcribing in the original language or translating to English. We find this approach is particularly effective at learning speech to text translation and outperforms the supervised SOTA on CoVoST2 to English translation zero-shot.\nWe hope Whisper’s high accuracy and ease of use will allow developers to add voice interfaces to a much wider set of applications. Check out the [paper(opens in a new window)](https://cdn.openai.com/papers/whisper.pdf), [model card(opens in a new window)](https://github.com/openai/whisper/blob/main/model-card.md), and [code(opens in a new window)](https://github.com/openai/whisper) to learn more details and to try out Whisper.\nReferences\nRelated articles\n[View all](/news/)",
  "body_base_path": "attachments/f5b4364f5d61/pages/",
  "body_markdown_path": "attachments/f5b4364f5d61/pages/archived-body.md",
  "local_html": "attachments/f5b4364f5d61/pages/archived-body.html",
  "body_source": "prior_capture"
}