douyin-extract-text

Extract captions and speech from TikTok videos using OCR and ASR.

18|4|Updated Jan 13, 2026
One-click install
npx skills add https://github.com/kaiye/skills --skill douyin-extract-text
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: douyin-extract-text
Source: https://github.com/kaiye/skills/tree/main/douyin-extract-text
Command: npx skills add https://github.com/kaiye/skills --skill douyin-extract-text

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires requests, rapidocr-onnxruntime, opencv-python, numpy, and includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill enables users to automatically extract text, captions, and speech content from TikTok videos, simplifying content analysis and repurposing.

Core Features & Use Cases

  • Video Text Extraction: Supports OCR and ASR modes to retrieve on-screen subtitles and spoken words from TikTok videos.
  • Multi-Modal Processing: Combines speech recognition and image-based text extraction for high accuracy.
  • Use Case: When analyzing popular TikTok videos for trending phrases, users can programmatically obtain accurate transcriptions and captions for content review.

Quick Start

Use the douyin-extract-text skill to extract captions from a TikTok video URL for further analysis.

Frequently Asked Questions about douyin-extract-text

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I extract captions and speech content from TikTok videos?

To extract captions and speech content from TikTok videos, you can use a combination of OCR and speech recognition techniques. This approach automates retrieving on-screen subtitles and spoken words for high accuracy transcription.

What is the best way to automate TikTok video text extraction for content analysis?

Automating TikTok video text extraction involves using multi-modal processing that combines speech recognition and image-based text extraction. This method programmatically obtains accurate transcriptions ideal for social media analysis and digital media archiving.

Do I need opencv-python and rapidocr-onnxruntime to process video captions?

Yes, you need opencv-python and rapidocr-onnxruntime to process video captions. These libraries, along with requests and numpy, provide the necessary environment for executing OCR and multi-modal speech recognition tasks.

Can I use OCR and ASR techniques for social media archiving?

Yes, you can use OCR and ASR techniques for social media archiving. Combining image-based text extraction and speech recognition allows you to accurately capture both on-screen subtitles and spoken words from video content.

How does multi-modal processing improve video text extraction accuracy?

Multi-modal processing improves video text extraction accuracy by combining speech recognition and image-based text extraction. This dual approach ensures both spoken words and on-screen subtitles are captured accurately from the video.