Wearable Speech Evaluation Using Listener Reaction Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing speech evaluation systems lack accuracy in determining evaluation values for speeches in multi-participant communication, as they do not adequately consider listener reactions occurring both during and after the speech.
Innovation Solution
A speech evaluation system utilizing wearable terminals with sound-collecting and acceleration sensors to detect speeches and calculate evaluation values based on listener reactions, including nodding motions, within a defined evaluation period that accounts for reactions both during and after the speech, ensuring accurate evaluation by setting the second timing appropriately.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the evaluation period is extended to include reactions after the speech ends, then the accuracy of speech evaluation is improved, but the calculation complexity increases
Solution Approach 1:
The patent pre-defines the evaluation period structure with clear start and end timing rules before actual speech evaluation occurs. The start timing is set to the speech start time or a predetermined time before, and the end timing is set to a predetermined time after speech end or when another speech begins. This preliminary temporal framework simplifies the calculation process while ensuring comprehensive capture of listener reactions both during and after speeches.
2Ease of manufacture
If the second timing is set to a predetermined time after speech end, then the calculation process is simplified, but listener reactions occurring between speeches may be incorrectly attributed
Solution Approach 1:
The patent implements a dynamic end timing determination mechanism that adapts to actual speech patterns. The end timing is not fixed but is determined by monitoring when the next speech begins. If another speech starts before the predetermined time elapses, the evaluation period ends at that point. This dynamic adjustment ensures accurate attribution of listener reactions to the correct speech while maintaining computational efficiency through the predetermined time threshold.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system accurately calculates evaluation values for each speech by considering listener reactions, enhancing the detection of important speeches and reducing calculation complexity, allowing for effective identification of crucial communication points.
Implementation Method 1
a sensor including at least a sound-collecting unit
Implementation Method 2
each of the plurality of wearable terminals being adapted to be worn by a respective one of the plurality of participants and including a sensor including at least a sound-collecting unit
Data Source
AI summary
A speech detection unit detects a speech in communication based on output values of microphones of a plurality of wearable terminals, and identifies a wearable terminal corresponding to the detected speech. A speech period detection unit detects for each speech a start timing of the speech and an end timing thereof. An evaluation-value calculation unit calculates, for each speech detected by the speech detection unit, an evaluation value for the speech based on an output value of an acceleration sensor of a wearable terminal other than the wearable terminal corresponding to the speech in an evaluation period from a first timing, which is at or later than the start timing of the speech and earlier than the end timing of the speech, to a second timing, which is later than the end timing of the speech.