Google Marketing Solutions has released Ariel, a new open-source tool designed to automate the translation and dubbing of video Ads into multiple languages. This solution aims to enhance the global reach of digital advertising by streamlining the localization process.
Core Features
Ariel offers several key features:
- Automated Dubbing: Generates high-quality dubbed versions of video ads in various target languages.
- Scalability: Handles large volumes of videos across diverse languages.
- User-Friendly Interface: Provides a straightforward API and user interface for ease of operation.
- Cost-Effective: Primary expenses limited to Gemini API and Text-To-Speech API calls, potentially reducing dubbing costs compared to traditional methods.
Technology Stack
Ariel utilizes a combination of AI and audio processing technologies:
- DEMUCS for audio source separation
- pyannote for speaker diarization
- faster-whisper and Gemini 1.5 Flash for speech-to-text conversion
- Gemini 1.5 Flash for translation
- Google Cloud Platform's Text-To-Speech or ElevenLabs API for generating speech in target languages
Language Support
The tool supports +40 languages, including major global languages and several regional dialects. Language coverage may vary depending on the underlying services used for speech-to-text, translation, and text-to-speech functions.
The tool is available for installation via pip and all details can be found here on Github.
