Visual Information Processing Allocation Between Mobile Device and Network
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Object recognition techniques on mobile devices consume significant battery power and computing resources, leading to latency, while offloading these tasks to a network can burden the network with excessive data transfer, increasing costs for users.
Innovation Solution
A mobile device system that determines whether to perform visual information processing locally or offload it to a network based on device and network conditions, using a processor to assess battery charge, network connectivity, and data transfer costs, and selects appropriate features and classification techniques to optimize processing between the device and network.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If object recognition is performed on a mobile device, then processing speed is improved, but battery power is consumed quickly
Solution Approach 1:
The patent segments the object recognition task into two parts: feature extraction is performed on the mobile device using computational geometry algorithms, while classification is performed on the network server. This segmentation allows the mobile device to perform lightweight processing locally while offloading the computationally intensive classification task to the network, thereby reducing battery consumption while maintaining processing speed.
Solution Approach 2:
The patent introduces an intermediary approach where the mobile device extracts features and transmits them to the network for classification. This intermediary process of feature extraction acts as a bridge between local processing capabilities and remote computational power, enabling the system to leverage both mobile device speed and network computational resources while minimizing battery consumption.
2Use of energy by moving object
If object recognition is offloaded to a network, then battery power is conserved, but data transfer burden increases
Solution Approach 1:
The patent extracts only the essential feature data from the original images and transmits this condensed information to the network for classification. By extracting and sending only the necessary feature representations rather than the complete images, the system significantly reduces the volume of data transferred to the network while still providing sufficient information for accurate object recognition.
3Measurement precision
If traditional object recognition techniques are used, then classification accuracy is achieved, but computing resources are excessive
Solution Approach 1:
The patent segments the computational workload by implementing simple computational geometry algorithms for feature extraction on the mobile device, while using machine learning classifiers on the network server. This segmentation allows the mobile device to use minimal computing resources for local feature extraction while the network handles the computationally intensive classification with sufficient accuracy.
Solution Approach 2:
The patent changes the approach by using computational geometry parameters and simple visual features for local processing, while relying on machine learning models for remote classification. This parameter change strategy allows the system to achieve high classification accuracy through the network while keeping mobile device computing requirements low.
Data Source
AI summary
An illustrative mobile device includes a data storage configured to at least temporarily store visual information and at least one processor that is configured to determine whether to request visual information processing from a network with which the mobile device may communicate. The processor is configured to determine a mobile device condition and a network condition. The processor determines a type of feature from the visual information to use for classification based on the determined mobile device and network conditions. The processor is configured to classify the visual information based on the determined type of feature and determine a confidence indicator based on the classification. The processor determines whether to request visual information processing from the network based on the determined confidence indicator.