NVIDIA ouvre l'accès de ses réseaux d'apprentissage profond (DNN) à l'industrie du transport pour accélérer le développement des véhicules autonomes
NVIDIA ouvre l'accès de ses réseaux d'apprentissage profond (DNN) à l'industrie du transport pour accélérer le développement des véhicules autonomes

L'industrie du transport va pouvoir profiter de l'accès aux réseaux d'apprentissage profond NVIDIA DRIVE™, afin de développer des véhicules autonomes via le registre de conteneurs NVIDIA GPU Cloud (NGC). Utilisé largement par les constructeurs d'automobiles, de camions et par les entreprises de robotaxis, NVIDIA DRIVE s'est imposé comme le standard en matière de développement de véhicules autonomes.

L'intelligence artificielle est essentielle au développement de véhicules autonomes sûrs car elle leur permet de percevoir ce qui les entoure et de réagir en temps réel. Ces véhicules sont équipés de douzaines de DNNs qui garantissent la précision de la perception, de la géolocalisation et de la planification des trajectoires.

« Les véhicules autonomes intelligents sont des véhicules software-defined qui doivent évoluer dans le monde grâce aux informations fournies par de nombreuses séries de données. En donnant l'accès aux développeurs à nos réseaux d'apprentissage profond, ainsi qu'à des outils d'apprentissage avancés pour optimiser ces DNN pour de multiples séries de données, nous facilitons le partage des connaissances entre différentes entreprises et différents pays, sans remettre en cause la propriété et la confidentialité des données. Finalement, ce sont tous les utilisateurs qui vont en profiter », explique Jensen Huang, fondateur et CEO de NVIDIA.


Le communiqué complet ci-dessous :

NVIDIA DRIVE has become a de facto standard for AV development, used broadly by automakers, truck manufacturers, robotaxi companies, software companies and universities. Now, NVIDIA is providing access of its pre-trained AI models and training code to AV developers. Using a suite of NVIDIA AI tools, the ecosystem can freely extend and customize the models to increase the robustness and capabilities of their self-driving systems.

“The AI autonomous vehicle is a software-defined vehicle required to operate around the world on a wide variety of datasets,” said Jensen Huang, founder and CEO of NVIDIA. “By providing AV developers access to our DNNs and the advanced learning tools to optimize them for multiple datasets, we're enabling shared learning across companies and countries, while maintaining data ownership and privacy. Ultimately, we are accelerating the reality of global autonomous vehicles.”

AI is central to the development of safe, self-driving cars — allowing them to perceive and react in real time to their surroundings for intelligent operation. At its core are dozens of DNNs that tackle redundant and diverse tasks, ensuring accurate perception, localization and path planning.

“NVIDIA leads the world in developing the deepest and broadest suite of DNNs and AI tools for the transportation industry,” said Luca De Ambroggi, senior research director of Artificial Intelligence at IHS Markit. “Making these algorithms available to others, along with the tools and workflow infrastructure to customize them, will help enable the deployment of safe autonomous transportation.”

NVIDIA has spent years developing and training DNNs that run on the NVIDIA DRIVE AGX™ platform, turning raw sensor data into a deep understanding of the world. These DNNs cover such tasks as traffic-light and sign detection, object detection (for vehicles, pedestrians, bicycles) and path perception, as well as gaze detection and gesture recognition inside the vehicle.

Advanced Learning Tools for AI Development

In addition to providing access to the DNNs, NVIDIA announced the availability of a suite of advanced tools so developers can customize and enhance NVIDIA's DNNs using their own datasets and target feature set. These tools allow the training of DNNs using active learning, federated learning and transfer learning:

  • Active learning improves model accuracy and reduces data collection costs by automating data selection using AI, rather than manual curation.
  • Federated learning enables companies to utilize datasets across countries and with other companies while maintaining data privacy and protecting their intellectual property.
  • Transfer learning gives DRIVE customers the ability to speed development of their perception software by leveraging NVIDIA's significant investment in AV development, then further developing these networks for their own applications and target capability.

By providing access to its AI models on NGC and introducing advanced training tools, NVIDIA strengthens its end-to-end platform for AV development and deployment.


Article original : NVIDIA