计算机科学
人工智能
模式识别(心理学)
情态动词
特征提取
转化(遗传学)
特征(语言学)
鉴定(生物学)
语音识别
算法
人工神经网络
计算机视觉
作者
Yang Liu,Alexandros Neophytou,Sunando Sengupta,Eric Sommerlade
出处
期刊:International Conference on Acoustics, Speech, and Signal Processing
日期:2021-06-06
卷期号:: 830-834
标识
DOI:10.1109/icassp39728.2021.9414779
摘要
Convolutional neural networks (CNNs) with log-mel spectrum features have shown promising results for acoustic scene classification tasks. However, the performance of these CNN based classifiers is still lacking as they do not generalise well for unknown environments. To address this issue, we introduce an acoustic spectrum transformation network where traditional log-mel spectrums are transformed into imagined visual features (IVF). The imagined visual features are learned by exploiting the relationship between audio and visual features present in video recordings. An auto-encoder is used to encode images as visual features and a transformation network learns how to generate imagined visual features from log-mel. Our model is trained on a large dataset of Youtube videos. We test our proposed method on the scene classification task of DCASE and ESC-50, where our method outperforms other spectrum features, especially for unseen environments.
科研通智能强力驱动
Strongly Powered by AbleSci AI