In-Vehicle Speech Recognition Using Tap Signals for Faster Commands
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech recognition systems in vehicles require multiple dialogues to confirm user speech and obtain additional information, causing user inconvenience.
Innovation Solution
A speech recognition system that maps and registers tap signals generated by tapping nearby objects in a vehicle to specific commands, allowing users to replace utterances with simple actions, thereby improving convenience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If speech recognition system uses multiple dialogues to confirm user speech and obtain additional information, then service accuracy is improved, but user convenience deteriorates
Solution Approach 1:
The patent segments the command input method into two distinct types: voice commands for complex instructions requiring confirmation, and tap signals for simple, direct commands. This segmentation allows the system to optimize for both accuracy (using dialogue confirmation for voice) and convenience (using direct tap execution), resolving the contradiction by applying different interaction modes to different command types.
Solution Approach 2:
Instead of requiring all commands to go through the traditional voice dialogue confirmation process, the patent inverts the approach by allowing tap signals to directly execute commands without confirmation dialogue. This inversion eliminates the unnecessary dialogue steps for tap-based commands while maintaining the ability to use dialogue confirmation for voice commands, thus improving convenience without sacrificing accuracy where needed.
2Measurement precision
If speech recognition system requires multiple dialogues for service confirmation, then service precision is improved, but response time deteriorates
Solution Approach 1:
The patent implements preliminary action by pre-mapping specific tap patterns (single tap, double tap, long press) to specific commands before execution. This pre-establishment of command mappings allows the system to execute tap-based commands immediately upon detection, eliminating the time-consuming dialogue confirmation process while maintaining service precision through the pre-defined mapping relationships.
Solution Approach 2:
The patent applies the skipping principle by allowing tap signals to bypass the dialogue confirmation step entirely and directly trigger command execution. This rushing through of the confirmation process for tap-based inputs significantly reduces response time while maintaining service precision through the pre-established tap-to-command mappings, effectively resolving the time-precision contradiction.
Data Source
AI summary
Provided are a speech recognition system and a method for providing a speech recognition service that may map and register a tap signal, generated by tapping an object around a user of a vehicle, to a specific command, and replace an utterance for the specific command with a simple action of tapping the nearby object, to improve a user convenience. A speech recognition system includes: a speech processing module configured to extract information from a voice signal of the user in a vehicle; a control module configured to generate a control signal for performing the control intended by the user; and a memory configured to map and store a tap signal and a command corresponding to the tap signal, wherein, when the tap signal is included in an audio signal input, the control module is configured to generate the control signal based on the command corresponding to the stored tap signal.