Voice Search Playlist Creation Using Speech Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The vast amount of content available on electronic devices, such as PDAs and cellular phones, makes it cumbersome to find specific songs, documents, or videos using simple menus, as users must sift through thousands or millions of options, wasting time and effort.
Innovation Solution
Implementing a speech recognition method that stores content with associated attribute values, generates likelihood values from speech input, and ranks them to efficiently access and play lists, allowing users to select content using voice commands and attributes like song titles or artist names.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If simple menus are used to search through content lists, then the device structure remains simple, but the time and effort required to find content increases significantly
Solution Approach 1:
The patent replaces the mechanical manual navigation through menus with an acoustic field-based speech recognition system. Users speak natural language queries which are converted to text and processed to retrieve content, eliminating the need to manually navigate through thousands of menu items and significantly reducing search time.
Solution Approach 2:
The patent introduces speech recognition technology as an intermediary between the user and the content database. Instead of directly interacting with menu structures, users communicate through speech which is processed by the recognition system to retrieve desired content, making the interaction more efficient and intuitive.
2Quantity of substance
If thousands of content items are stored on a single device, then content availability increases, but the complexity of content selection increases
Solution Approach 1:
The patent replaces complex manual content selection processes with speech-based natural language processing. Users can query content using spoken descriptions rather than navigating complex menu hierarchies, making content selection as simple as speaking and thereby reducing operational complexity despite large content volumes.
Solution Approach 2:
The speech recognition system serves as a universal interface for accessing all types of content (music, videos, documents, etc.) regardless of their category or storage location. This single multi-functional interface replaces the need for separate access methods for different content types, simplifying the overall selection process.
3Use of energy by moving object
If manual navigation through content lists is used, then the system requires minimal processing power, but user productivity decreases
Solution Approach 1:
The patent replaces energy-efficient but slow manual navigation with a speech recognition system that uses acoustic processing and natural language understanding. Although this requires more processing power, it dramatically improves productivity by enabling users to access content through intuitive spoken commands rather than tedious manual searching.
Data Source
AI summary
Embodiments of the present invention improve content selection systems and methods using speech recognition. In one embodiment, the present invention includes a speech recognition method comprising storing content on an electronic device, wherein the content is associated with a plurality of content attribute values, adding the content attribute values to a first recognition set of a speech recognizer, receiving a speech input signal in said speech recognizer, generating a plurality of likelihood values in response to the speech input signal, wherein each likelihood value is associated with one content attribute value in the recognition set; and accessing the stored content based on the likelihood values.