See What's NEW

Industry Solutions

Automotive

automotive-img

Automotive

Magic Data provides high quality training and test data for your AI models in many scenarios of intelligent cockpit and autonomous driving to continuously improve the performance of your AI applications.

With industry Know-How as the key feature, we are embracing this new challenge with you and committed to creating more opportunities in the wave of digitalization, automation, and AI.

Contact Sales

Scenarios

0-img

Navigation

Device wake-up, navigation command recognition, POI recognition

1-img

Device Control in or out of the Vehicle

Car window, A/C, seat, lighting system, vehicle door, windscreen wiper

2-img

Virtual Assistant

Information query, travel arrangement, phone call, entertainment

3-img

DMS

Fatigue detection, distraction detection, poor driver behaviour detection

4-img

OMS

Seat belt reminder, dangerous behavior detection, Reminder for carrying baby in vehicle

5-img

Autonomous Vehicle/Advanced Driving Assistance System

Street view semantic recognition, road markings/traffic signs recognition, anomalous object detection

Challenge

In-vehicle scenario
Limited data of fatigue, distraction, and poor driver behavior
Unable to correctly understand commands and queries
Strong or dim light due to environment and weather changes
Stiff and unnatural response
Limited data of emergency and abnormal scenarios

Annotator® AI-Assisted Annotation Platform

Audio Annotation Text Annotation Image Annotation
  • Navigation - Navigation command annotation
  • Device control in or out of the vehicle - Device command annotation
  • Virtual assistant - Voice assistant command and query annotation
annotator-img
  • Navigation - Speech command generalization
  • Device control in or out of the vehicle - Speech command generalization
  • Virtual Assistant - Command and query generalization
annotator-img
  • DMS - Keypoint annotation for facial recognition
  • OMS - In-vehicle bounding box annotation for passenger and object recognition
  • CV - Bounding box annotation for out of the vehicle entity annotation, street view semantic segmentation annotation
annotator-img

MD Dataset Portfolio

Speech Recognition
Text-to-Speech
Natural Language Understanding
OCR

Contact us for data collection and annotation service

annotator-serve-img

Related Datasets

Multi-Emotional Natural Speech Dataset

Magic Data has newly introduced the "Multi-Emotional Natural Speech Dataset", comprising various datasets designed to enhance expressiveness and naturalness in speech technology, enabling intelligent devices to exhibit a wide range of emotional expressions. This dataset significantly enhances the emotional expressiveness of large speech models. By leveraging our dataset, the expressiveness and emotional authenticity of large speech models can be greatly improved.
Play Audio

MDT-LF001 Changsha Dialect Lexicon

MDT-NF023 Mandarin Chinese Human–Computer Interaction Text Corpus

MDT-LE001 Shanxi Dialect Lexicon

MDT-NG031 Vehicle command control Text Corpus

MDT-AI101 Spanish Duplex Conversation Training Dataset

Preserving features such as tonal jumps, spontaneous interruptions, and collaborative speech in fast-paced native conversations, this dataset uses independent channel recording for precise voice separation. Combined with multi-speaker labeling and scenario classification, it provides a solid training foundation for AI models to manage diverse speech rates and regional linguistic variations in Spanish.

Contact us for the best practices

Get started today

TOP
Talk to Magic Data