Remote Machine Learning Model Proxy for Resource Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in efficiently and effectively executing machine-learned models remotely, particularly in maintaining control over proprietary models, ensuring compatibility with various user devices, and optimizing resource usage, especially when user devices have limited computing resources.
Innovation Solution
The implementation of a remote execution process that detects wireless signals from a host device, establishes a communication connection, and generates a proxy to the machine-learned model, allowing user devices to access and utilize the model seamlessly, thereby avoiding the need for local resource overload and ensuring compatibility with different application versions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If machine-learned models are executed locally on user devices, then processing speed and responsiveness are improved, but device resource consumption and complexity increase
Solution Approach 1:
The patent introduces a proxy as an intermediary component that resides on the user device but communicates with and manages access to the machine-learned model on the remote host device. This proxy handles model execution requests, manages data transmission, and coordinates inference operations, thereby enabling fast local-like access while the actual computationally intensive model processing occurs remotely on the host device with sufficient resources.
2Reliability
If proprietary machine-learned models are hosted remotely, then control and security are improved, but accessibility and ease of operation deteriorate
Solution Approach 1:
The patent creates a proxy copy or representation of the machine-learned model interface on the user device. This proxy replicates the model's input-output behavior and API interface, allowing applications to interact with the model as if it were locally present, while the actual proprietary model remains securely hosted on the remote host device. This copying approach maintains both security and ease of access.
3Adaptability or versatility
If machine-learned models are distributed to multiple user devices, then compatibility and ease of operation are improved, but model security and control deteriorate
Solution Approach 1:
The patent extracts the machine-learned model from the user devices and relocates it exclusively to the remote host device. By taking the model out of the distributed environment and centralizing it on the host, the system maintains model security and control while still providing access to multiple user devices through the proxy mechanism. This extraction eliminates security risks associated with model distribution while preserving compatibility.
4Use of energy by moving object
If machine-learned models are executed on host devices, then resource efficiency is improved, but communication overhead and loss of time increase
Solution Approach 1:
The patent implements preliminary actions by establishing the proxy connection and pre-configuring the communication channel between the user device and host device before actual model inference is needed. The proxy is set up in advance to manage data formatting, transmission protocols, and buffer management, thereby reducing communication overhead and latency during actual model execution. This preliminary setup optimizes the subsequent interaction efficiency.
Data Source
AI summary
In an exemplary process for remote execution of machine-learned models, one or more signals from a second electronic device is detected by a first electronic device. The second electronic device includes a machine-learned model associated with an application implemented on the first electronic device. Based on the one or more signals, a communication connection is established with the second electronic device and a proxy to the machine-learned model is generated. Input data is obtained via a sensor of the first electronic device. A representation of the input data is sent to the second electronic device via the proxy and the established communication connection. The representation of the input data is processed through the machine-learned model to generate an output. A result derived from the output is received via the communication connection and a representation of the result is outputted.