Case Study for Speech Data Collection

From：Nexdata Date： 2024-04-02

The client has excelled in the voice-related context awareness/dialogue middleware engine and voice handler for the automotive industry. Their efforts to provide voice recognition and a conversational, natural dialogue engine for automobiles and digital devices will revolutionize how we communicate with man-made machines.

Challenge

The client has a new innovative patented engine which means they need the support of hundreds of hours of speech data collection in various languages and locations around the world of all possible terms, accents, and phrases that would use to communicate in the vehicle.

Solution

To provide high-quality speech data, Nexdata visited 11 countries with our expert linguists to record over 30,000 colloquial sentences.

Results

We deliver the required utterances in 11 languages within stringent timelines and with the required quality, with a sentence accuracy rate of 97%. The client’s models can understand multiple languages and truly change how we drive and live.

Case Study for Speech Data Collection

Recent

Behavior Detection Data: Enhancing Systems through Human Behavior Analysis

Text-to-Speech (TTS) Data: Fueling the Future of Synthetic Voices

Human Voice Datasets: A Key Resource for Speech Technology Development

Previous

Multi-Sensor Fusion Perception in Autonomous Driving

Next

Case study in autonomous driving