On-Device AI Product Recognition With Power-Managed Frame Rates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional online product purchase systems are cumbersome and inconvenient, requiring users to search for products online and manually gather information, and offline product searches are inefficient due to difficulties in real-time AI recognition and high data transmission requirements, leading to delayed responses and reduced recognition rates for new products.
Innovation Solution
A mobile artificial neural network device equipped with a camera and AI recognition model that performs real-time product recognition, displaying information through augmented reality, with features like frame rate adjustment, battery conservation, and communication with servers for additional data, using lightened AI models to minimize computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-definition video is transmitted to the Internet server in real time for AI recognition, then product information can be recognized, but the amount of data transmission increases significantly and response speed is delayed
Solution Approach 1:
The patent segments the AI recognition process by deploying a lightweight AI model directly on the mobile terminal rather than relying on server-based processing. This divides the system into local inference capabilities and selective cloud communication, enabling real-time recognition without continuous high-definition video transmission.
Solution Approach 2:
The patent extracts the essential AI recognition functionality from the server environment and embeds it directly into the mobile terminal. By taking out the core inference engine and placing it locally, the system eliminates the need for continuous video streaming while maintaining recognition capabilities.
2Adaptability or versatility
If the AI recognition model is stored in the Internet server, then recognition requests can be processed centrally, but users cannot monopolize the model in real time and response speed is delayed depending on the number of users
Solution Approach 1:
The patent applies local quality by implementing AI recognition capabilities directly on the user's mobile terminal rather than relying on remote server processing. This ensures that each user has dedicated, consistent access to the recognition model with guaranteed response times independent of server load or other users' activities.
3Measurement precision
If the AI recognition model learns new products by itself, then recognition rate for new products can be improved, but considerable power consumption and computational amount are required
Solution Approach 1:
The patent applies preliminary action by pre-training the AI model on a comprehensive dataset of products before deployment on the mobile terminal. The model is prepared in advance with extensive product knowledge, enabling it to recognize new products without requiring energy-intensive learning operations during actual use.
Data Source
AI summary
A mobile artificial intelligence (AI) device includes a camera configured to capture a video of a product at a first frame rate, and an AI recognition model configured to recognize product information from the captured video on the mobile AI device. A processor is configured to operate the AI recognition model at a second frame rate, which is selectively adjustable to manage power consumption. A display is configured to concurrently present the video at the first frame rate and the recognized product information at the second frame rate.