Ru2En Interpreter
An iOS Russian–English voice interpreter for conversations that need to move naturally in both directions.
Problem
Real-time bilingual conversations are difficult when people must repeatedly choose a language, manage turns, and switch between listening and speaking.
Solution
A voice interpreter that supports Russian–English translation in both directions while detecting language and conversational turns automatically.
Architecture
The iOS prototype combines speech recognition, automatic language and turn detection, translation, text-to-speech, conversation history, and AirPods audio routing.
Technical decisions
- Treat each conversation as bidirectional rather than a single-direction translation task.
- Use automatic language and turn detection to reduce interaction overhead.
- Support AirPods routing for more natural hands-free use.
- Keep conversation history available for context and review.
Challenges
- Balancing automatic behavior with the ambiguity of live conversation.
- Keeping the audio experience coherent across recognition, translation, and speech output.
Results
Public performance metrics are not available for this project.
Current status
Prototype.