SoundHound ports generative AI to Nvidia Drive, sees Stellantis launch
Cette publication existe aussi en Français
SoundHound AI has developed an in-vehicle voice assistant that uses a large language model (LLM) running on the Nvidia DRIVE platform and is rolling out its technology with Stellantis in Japan.
SoundHound’s work with Nvidia will allow it to significantly expand the number of places and situations that generative AI can be deployed without the need for a cloud connection to run the LLM in a datacentre.
Among a range of use cases, the SoundHound Vehicle Intelligence provides information directly from the car manual and other relevant data sources using natural speech using vehicle specific retrieval-augmented generative (RAG) AI. The next generation Thor Drive processor will combine ARM CPUs with the new Blackwell GPU core.
The previous voice assistant with integrated ChatGPT will be the first to go into vehicles in Japan with European car maker Stellantis. SoundHound Chat AI Automotive launched in April 2023 and will be available in Stellantis DS Automobiles in Japan starting this month.
At the beginning of March 2024, DS Automobiles became the first automaker in the world to go into full production with SoundHound Chat AI, with an initial rollout in 13 languages across 18 countries. This assistant – which DS Automobiles has named Iris – will allow drivers and passengers to use hands-free voice control to unlock a vast range of information and updates.
The Vehicle Intelligence tool understands a verbal request and seamlessly provides answers – including settings, safety, troubleshooting, and vehicle maintenance – without the need for a cumbersome physical document. Examples include:
“I see a flashing light that looks like a car battery and I’m not sure what that means?”
“What does the ‘auto hold’ button do?”
“How do I use that feature that lets me drive hands-free safely?”
In addition to Vehicle Intelligence, users can ask more general questions that can help drivers plan a trip or a vacation, such as:
“Where are the best locations to take photographs on the Pacific Coastal Highway?”
“Which wineries offer riesling in Carmel Valley?”
“What kinds of dishes count as Californian cuisine?”
This technology opens up opportunities for car makers looking to give drivers rapid voice-enabled access to LLM capabilities with the added benefit of greater privacy, flexibility, and lower operating costs.
“Together with Nvidia, we’re marrying the incredible capabilities of generative AI with all the advantages of edge computing,” said Mike Zagorsek, COO of SoundHound AI. “The net result is a fast and private voice experience with seamless results. And with this new level of adaptability, the possibilities are endless.”
“We’re working with innovative partners like SoundHound to bring generative AI and accelerated compute into the car – enhancing the occupant experience and bringing greater safety behind the wheel,” said Rishi Dhall, Vice President of Automotive at NVIDIA. “SoundHound’s in-vehicle voice interface, powered by NVIDIA DRIVE, can provide drivers with fast, accurate information, even when there’s no connection.”
SoundHound AI offers a selection of on-device edge voice solutions that allow automakers to grant their end-users greater levels of privacy by keeping data stored locally.
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
