Speech Recognition Text Delivery via Remote Server
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Historically, mobile computing devices have been limited in their ability to perform accurate speech recognition due to insufficient computing power and limited network connectivity, and existing solutions require companion applications to be installed on target devices for speech-derived text to be usable.
Innovation Solution
A system and method where a speech recognition module on a mobile device generates text from speech input and transmits it to a remote storage device, which makes the text accessible via a network to any computing device with a web browser, enabling text processing and display across multiple devices without the need for specific companion applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If speech recognition is performed on mobile devices with limited computing power, then the device can operate independently, but the accuracy and speed of speech recognition deteriorates
Solution Approach 1:
The speech recognition system is segmented into two parts: the mobile device captures and transmits audio data, while the actual speech recognition processing is performed on a remote server with sufficient computing power. This segmentation allows the mobile device to remain lightweight while achieving accurate speech recognition through cloud-based processing.
Solution Approach 2:
A remote server acts as an intermediary between the mobile device and the speech recognition process. The server receives audio data from the mobile device, performs the computationally intensive speech recognition, and returns the recognized text. This intermediary enables accurate speech recognition without requiring the mobile device to have sufficient computing power.
2Measurement precision
If speech recognition processing is performed remotely on another computing device, then speech recognition accuracy improves, but network connectivity requirements increase and device compatibility decreases
Solution Approach 1:
The system uses universal web technologies (HTML, CSS, JavaScript) to create the interface, making it accessible from any device with a web browser. The speech recognition service can be accessed from mobile devices, tablets, desktops, or any other computing device with internet connectivity, without requiring device-specific applications or software installation.
Solution Approach 2:
A web server acts as a universal intermediary that communicates with any device through standard web protocols. The server provides speech recognition services through a web interface that can be accessed from any device with a browser, ensuring broad compatibility while maintaining high recognition accuracy through remote processing.
3Reliability
If a companion application is installed on the target computing device to receive speech-derived text, then the text can be delivered to the device, but the complexity of the system increases and ease of use decreases
Solution Approach 1:
The system allows users to access their speech-derived text directly through a web browser without requiring installation or configuration of companion applications. The web interface automatically retrieves and displays recognized text, and users can copy it to their clipboard with a single click, making the system self-service oriented and eliminating the need for additional software on target devices.
Solution Approach 2:
The system uses the web browser's native clipboard functionality to transfer recognized text from the server to the user's device. Instead of requiring a companion application to handle text delivery, the system leverages the universal copy-paste mechanism already present in all web browsers, simplifying the user experience while maintaining reliable text delivery.
4Measurement precision
If mobile devices transmit audio data for speech recognition, then speech recognition accuracy improves, but network bandwidth requirements increase
Solution Approach 1:
The system transmits only the necessary audio data for speech recognition to the remote server, avoiding transmission of excessive information. By sending only the raw audio recording (without additional metadata or processing), the system minimizes network bandwidth usage while providing the server with sufficient data to perform accurate speech recognition.
Data Source
AI summary
Aspects of the disclosure provide systems and methods for facilitating dictation. Speech input may be provided to an audio input device of a computing device. A speech recognition engine at the computing device may obtain text corresponding to the speech input. The computing device may transmit the text to a remotely-located storage device. A login webpage that includes a session identifier may be accessed from a target computing device also located remotely relative to the storage device. The session identifier may be transmitted to the storage device and, in response, a text display webpage may be received at the target computing device. The text display webpage may include the speech-derived text and may be configured to automatically copy the text to a copy buffer of the target computing device. The speech-derived text may also be provided to native applications at target computing devices or NLU engines for natural language processing.