Convolutional Neural Network for Multi-Intent Speech Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current speech recognition technologies face limitations in accurately recognizing multiple intent words within spoken speech, which hinders the effective operation of home appliances and other devices controlled by voice commands.
Innovation Solution
A convolutional neural network structure is employed to preprocess and analyze speech data, converting it into text and embedding data, then using convolution operations and max pooling to extract feature maps that are concatenated and processed to identify multiple intent words, enabling accurate recognition and classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional speech recognition methods are used, then the system is simple to implement, but the accuracy of recognizing multiple intent words is limited
Solution Approach 1:
The speech recognition system is divided into multiple specialized components: an intent word detection unit that identifies potential intent words, an attention mechanism that selectively focuses on relevant parts of speech, and a classification unit that determines the specific intent. This segmentation allows each component to specialize in a specific task, improving overall accuracy while managing complexity through modular design.
Solution Approach 2:
An attention mechanism is introduced as an intermediary between the speech input and the classification unit. This attention mechanism selectively weights different parts of the speech input, allowing the system to focus computational resources on the most relevant portions for intent recognition, thereby improving accuracy without proportionally increasing overall system complexity.
2Loss of information
If multiple intent words are recognized from speech, then the understanding of user intent is improved, but the processing time increases
Solution Approach 1:
The system performs preliminary processing by detecting potential intent words and computing attention weights before final classification. The attention mechanism pre-identifies which parts of the speech are most relevant to intent recognition, allowing the classification unit to focus only on these critical regions, thereby reducing the effective processing time while maintaining complete intent understanding.
Solution Approach 2:
The attention mechanism applies partial processing by selectively focusing computational resources only on the most relevant portions of the speech input rather than processing the entire speech signal uniformly. This allows the system to achieve complete intent understanding with reduced computational effort and shorter processing time by ignoring less relevant speech segments.
Data Source
AI summary
The present disclosure relates to a convolutional-neural-network structure for acquiring intent words, and a speech recognition device and method using the network. the method includes receiving input data generated from speech, performing convolution on the input data and N3 filters each having N2 channels, and acquiring a feature map having N4 pieces of data for each channel, applying max pooling to the N4 pieces of data to acquire a representative value, and acquiring a feature map having N2 pieces of data for each filter, performing concatenation on the feature maps for the respective filters, and acquiring one feature map of an N3×N2 matrix, performing convolution on the feature map of the N3×N2 matrix and a filter of a 1×N3 matrix, and acquiring a feature map of a 1×N2 matrix; and inputting the feature map of the 1×N2 matrix into an artificial neural network, and acquiring at least one intent word.