|
|
|
|
AVS generative GPT audio-visual interactive edge smart box |
|
|
The generative GPT audio-visual interactive edge smart box is a small system developed by AVS, featuring the AI860 chip and integrated with a Raspberry Pi single-board computer. This setup creates an "AI interactive audio-visual small front desk," which handles front-end tasks such as vehicle, parking space, and license plate recognition integration, as well as audio-visual interaction. It also converts natural voice inquiries from visitors into text for easier subsequent processing.
The text response analysis is conducted by the company's local GPT mini-station middleware, with Nvidia Nano or AI PC running the LLAMA GPT lightweight model to select responses (filtering or adjusting content that should not be displayed) and provide high-quality audio-visual intelligent interaction for exclusive domains such as fire prevention, disaster prevention, and other public or emergency services.
When expanding the GPT application scope, the middleware can also connect to cloud services from OpenAI or Google. With this middleware, it can serve and manage hundreds of sets of AVS interactive smart front desks, creating the most efficient and cost-effective high-quality intelligent AI service system for entire communities or groups (enterprises, governments, schools, various industries, etc.).
Example applications:
"I'm not familiar here... Where can I park my car?"
"How are the fees calculated?"
"Can I dine here or use the restroom?"
"Is there a convenience store nearby to buy things?"
"What time is it now?"
"How's the weather?"
"Oh no, there's a robbery... a fight broke out!"
"And someone is setting fire to a car... help!"
|
|
|
|
|
|
|
|
|