SenseVoice

Visit Tool

SenseVoice is an Audio & Music tool that provides a multilingual speech foundation model for understanding voice. It offers high-accuracy automatic speech recognition, spoken language identification, speech emotion recognition, and audio event detection.

Claim this tool

No Views Yet

At a glance

Pricing

Open Source

Free tier

Yes

API

Yes

Skill level

Technical

About

What is SenseVoice?

SenseVoice is a comprehensive speech foundation model designed for multilingual voice understanding. It integrates multiple speech understanding capabilities, including automatic speech recognition (ASR) for over 50 languages, spoken language identification (LID), speech emotion recognition (SER), and audio event detection (AED). Trained with over 400,000 hours of data, SenseVoice-Small boasts exceptionally low inference latency, processing 10 seconds of audio in just 70ms, making it 15 times faster than Whisper-Large. It also provides convenient finetuning scripts, service deployment pipelines supporting various client-side languages, and export features for ONNX and libtorch, making it suitable for both research and practical applications.

Best used for

Ideal for developers and researchers who need to integrate advanced speech understanding into their applications, analyze audio for emotional content, and detect specific sound events. Especially valuable for building multilingual voice-enabled systems with high accuracy and low latency.

Common actions

transcribe speech

identify language

detect emotions

recognize audio events

process audio efficiently

collaborationlow-code/no-codedeepfakeautomated workflowopen-sourceworkflowsface swappinggithub copilot"AI Agents"

Capabilities

Key features

Multilingual speech recognition
Spoken language identification
Speech emotion recognition
Audio event detection
Low-latency inference
Finetuning scripts
Service deployment

Target Audience

developersresearchersmachine learning engineers

Integrations

Not yet documented

Pricing & Plans

Open Source

Free

FAQs

What languages does SenseVoice support for speech recognition?

SenseVoice supports over 50 languages for multilingual speech recognition, including specific high-precision capabilities for Mandarin, Cantonese, English, Japanese, and Korean. It can also automatically detect the language of the input audio.

How fast is SenseVoice for audio processing?

The SenseVoice-Small model utilizes a non-autoregressive end-to-end framework, resulting in exceptionally low inference latency. It requires only 70ms to process 10 seconds of audio, making it 15 times faster than Whisper-Large.

Can SenseVoice be deployed in a Docker environment?

Yes, SenseVoice provides Docker support, allowing users to build and run the model using Docker for simplified setup, reproducibility, and support for both CPU and GPU inference. Docker Compose is also available for easier deployment.

Trending

Subcategories trending in Content & Design

Image Generation AI Writing Assistants Video Generation Photo Editing Graphic Design Video Editing

Trending

Explore

Browse AI tools by category

Content & Design Productivity & Business Coding & Development AI Agents & Automation Research & Education Wellness & Lifestyle Career Development Marketing & Growth Data & Analytics Customer Support & CX Finance E-commerce