Application and Research of Intelligent Speech Recognition Technology in Barrier-Free Human-Computer Interaction

Aiming at the problems of low efficiency and insufficient recognition accuracy of disabled users in the existing barrier-free human-computer interaction system, intelligent speech recognition technology is introduced as an improvement tool to improve the interactive response speed and operation convenience of the system. First, based on the deep learning model combining Convolutional Neural Network ( CNN ) and Long Short-Term Memory Network ( LSTM), audio samples suited to barrier-free interaction scenarios are curated, and speech data with real scene interference such as environmental noise and background music are collected to construct a barrier-free dataset specific to speech recognition. Then, in the low layer, the convolutional neural network is used to capture low-level audio features such as spectrum images and time domain. In the high layer, the model inputs the feature vector output by the convolution layer into the long short-term memory network (LSTM) to capture the temporal dependency of speech and adapt to long instructions and complex sentences. The attention mechanism is introduced through the hidden layer to enhance the focus on key features. The features are classified and output through the fully connected layer to achieve efficient speech feature extraction and dynamic learning in complex environments. Finally, its applicability and stability are verified in actual barrier-free interaction devices in quiet, noisy and background music environments. After the introduction of intelligent speech recognition technology, in the background music environment, the recognition accuracy fluctuated slightly around the expected value of 94.8%, with the maximum value close to 95.5% and the minimum value about 93.9%. The application of intelligent speech recognition technology in barrier-free human-computer interaction has effectively improved the quality of interaction and ease of use, offering a novel pathway for advancing barrier-free technologies in the future.