New York, April 3, 2025 - OORT, the trailblazer in the decentralized AI era, has announced the launch of its first batch of community-contributed in-car voice command datasets on Kaggle, the data science competition and collaboration platform under Google. This milestone marks a major step in enabling AI researchers and developers to improve voice command recognition in automotive systems using decentralized data collection.
Centralized in-car voice data collection always faces challenges like limited diversity, high costs, privacy concerns, and bias in validation, leading to suboptimal AI performance. Scalability issues and regulatory hurdles further restrict dataset quality and accessibility.
The dataset was carefully curated and filtered from a pool of over 85,000 data points divided into various categories, such as In-Car Command, Improving Speech Tech with Real Data, Smart Home Assistants, Book Titles Listing Audio, and more. These data points were gathered through the OORT DataHub's decentralized data marketplace, where contributors from around the world participated in more than 100 distributed tasks, contributing to the dataset's richness and diversity.
Over 1,000 data points, specifically from the in-car command category, are now listed on the Kaggle platform. This targeted dataset is designed to facilitate and accelerate the development and refinement of machine learning models that underpin voice command interfaces in automotive settings.
Decentralized approaches like OORT DataHub offer a scalable, transparent, and secure alternative for AI-driven voice recognition. The data underwent rigorous review and validation using OORT’s patented Proof-of-Honesty (PoH) consensus algorithm, ensuring accuracy and integrity before being categorized into five key in-car voice commands:
Each command features over 200 recorded samples, providing a rich and diverse dataset for training machine-learning models in real-world driving environments. As voice command systems become increasingly integral to modern vehicles, enhancing their accuracy and responsiveness—even in noisy conditions—is critical for improving safety and user experience.
The dataset is now publicly available on Kaggle, enabling AI practitioners, researchers, and enterprises to train, fine-tune, and evaluate machine learning models for automotive voice recognition. With an industry-driven structure, it is designed for seamless integration into AI pipelines, allowing further expansion with additional samples as needed.
This initiative underscores OORT’s commitment to decentralization, data privacy, and AI innovation, providing enterprises and developers with high-quality, real-world datasets to drive industry advancements.
For more details and access to the dataset, visit https://www.kaggle.com/datasets/oortdatahub/in-car-command
About OORT
OORT is the trailblazer in the decentralized AI era, offering a trustless infrastructure built for the future of AI. Powered by the Olympus protocol, OORT provides enterprises and individuals with a suite of decentralized AI products: OORT Storage, OORT DataHub (B2C, B2B), and OORT Compute (coming soon). OORT has raised $10 million from prominent investors to date, including Taisu Venture, Red Beard Venture, Sanctor Capital, and has received grants from Microsoft and Google.
Official website: https://www.oortech.com/
Follow us on: X | LinkedIn | Telegram | Discord