With the proliferation of mobile devices, consumers expect to stay connected, no matter where they are — especially in their car. Voice recognition in cars, however, continues to be the leading complaint among new vehicle owners. Global automotive manufacturers recognize the demand for better connectivity but face added complexities associated with localizing in-car systems for multiple languages.
Data collection & localization challenges:
- Training in-car systems requires a large volume of audio and translation data — difficult for in-house engineering teams to plan for, collect, and manage.
- Speech needs to be collected in various driving environments to account for real-world driving scenarios.
- Engineers often lack the language background needed for accurate speech data collection.
Because of these challenges, OEMs often outsource efforts to firms with deep linguistic backgrounds like Appen. A top global OEM has worked with Appen for over a decade to train its voice recognition systems for more than 20 languages. Appen offers a full-service approach for localization, data collection, in-car testing and validation, and linguistic consulting.
Appen worked closely with the OEM’s engineers to:
- Improve the accuracy of the synthesized voice that responds to driver commands
- Comprehensively test various in-car setups across languages
- Collect speech in simulated driving environments
- Ensure translations are consistent
- Provide unique value by supporting the OEM’s large-scale localization efforts
Working with Appen allowed the OEM’s engineers to focus on the core development work needed to develop leading in-car systems. Appen’s years of experience in the automotive industry allowed them to quickly develop robust data collection and validation programs that consistently meet their needs. Thanks to Appen’s comprehensive solution, the OEM has expanded into new markets, confident that its voice recognition system has been trained by high-quality data, backed by extensive linguistic capabilities.