Create

Sign in to ReadmeX

or

How we handle your data: Privacy Policy

ReadmeX

Your AI reading companion

What would you like to catch up on?

Pick a briefing, or ask your own question.

Your community and people briefings stay personal to you.

Story

iFlytek launches SparkAudio 1.0 voice foundation model

AI summary

iFlytek announced SparkAudio 1.0, a voice foundation model combining machine translation, speech recognition, semantic understanding, real-time speaker separation, environment understanding, audio understanding, and emotion recognition. The model is available in iFlytek Input Method 16.0, iFlytek Hearing, iFlytek Office Notebook, and iFlytek Translator, with Chinese-English switching and real-time speaker recognition. iFlytek also launched the iFlytek AI Recording Ring, a device with a 4+1 microphone array and Wi-Fi 6; its claimed accuracy and transcription-speed figures were not independently verified in the reports.

In-depth insight

The launch expands AI speech systems beyond transcription toward unified understanding of speakers, surroundings, audio, and emotion, while the reported product deployments make the capabilities available in everyday iFlytek tools.

科大讯飞在今日下午的全球1024开发者节上,把星火家族的第一块语音基石正式亮了出来——星火语音基座大模型 SparkAudio1.0。它把机器翻译、语音识别、语义理解这几样本领拧进同一个模型里,让基础语音能力整体往上抬了一截。

在官方演示里,SparkAudio1.0借一段电影片段秀出了它的"耳朵"有多尖:不仅能精准分辨声音内容、背景和说话人的性别,还能对场景做出准确判断。官方称,它在实时说话人分离、环境理解、音频理解和情绪识别上都"表现卓越",正把人工智能往真实物理世界的应用推进一步。

image.png

这块新基石已经落进讯飞的多款产品里。讯飞输入法16.0、讯飞听见、讯飞办公本、讯飞翻译机全面上线,支持中英文任意切换和说话人实时识别——也就是说,你边说边在两种语言之间跳转、身旁又有几个人先后开口,它都能跟得上、分得清。

配套硬件也一并端了出来。科大讯飞发布"讯飞 AI 录音环",搭载4+1环麦阵列、走 Wi-Fi6传输,宣称综合识别准确率超过96%,一段2小时的录音转写只要2分钟就能跑完。

#科大讯飞#SparkAudio 1.0#讯飞输入法 16.0#讯飞听见#讯飞办公本#讯飞翻译机#讯飞 AI 录音环

Original · 1 min readRead the original
Group chat

No comments yet. Start the conversation.