Tencent Cloud ASR Skill

Purpose

Provide capabilities for voice-to-text conversion and voice dialogue development using Tencent Cloud ASR API.

When to Use This Skill

Use this skill when:
- User asks to convert audio to text using Tencent Cloud ASR
- User wants to implement voice dialogue or voice assistant
- User needs to process audio files with speech recognition
- User wants to integrate Tencent Cloud ASR into their application

How to Use

Prerequisites

1. Get Tencent Cloud Credentials:
   - SecretId and SecretKey from Tencent Cloud Console
   - Enable ASR service at Tencent Cloud ASR

2. Install Dependencies:
   bash pip install tencentcloud-sdk-python-core # or npm install tencentcloud-sdk-nodejs

Available Scripts

1. Audio Transcription (scripts/asr_client.py)

Use scripts/asr_client.py for batch audio file transcription.

Python Usage:

from scripts.asr_client import TencentASR

# Initialize with your credentials
asr = TencentASR(secret_id="your_secret_id", secret_key="your_secret_key")

# Transcribe audio file
result = asr.recognize("path/to/audio.wav", engine_type="16k")
print(result)

2. Real-time ASR (scripts/realtime_asr.py)

Use scripts/realtime_asr.py for real-time speech recognition.

Features:
- Stream audio input
- Real-time text output
- Support for dialogue mode

Configuration Options

Voice Dialogue Development

For building voice dialogue systems:

1. ASR Module - Convert speech to text
2. NLP Module - Process text intent (can use Tencent Cloud TI-ONE or other NLP services)
3. TTS Module - Convert response to speech (Tencent Cloud TTS)
4. Dialog Manager - Handle conversation flow

Reference Documentation

See references/tencent_asr_api.md for detailed API documentation.

Supported Audio Formats

• PCM/WAV (recommended)
• MP3
• AMR
• AAC
• OGG

Rate Limits

• Synchronous recognition: 60 requests/minute
• File size limit: 60MB
• Audio duration limit: 60 seconds (for sync)