School of Instrument and Electronics, North University of China, Taiyuan, 030051, China.
Key Laboratory of Instrumentation Science and Dynamic Measurement (North University of China), Ministry of Education, Taiyuan, 030051, China.
Sci Rep. 2021 Apr 7;11(1):7648. doi: 10.1038/s41598-021-87399-1.
Due to the audio information of different types of vehicle models are distinct, the vehicle information can be identified by the audio signal of vehicle accurately. In real life, in order to determine the type of vehicle, we do not need to obtain the visual information of vehicles and just need to obtain the audio information. In this paper, we extract and stitching different features from different aspects: Mel frequency cepstrum coefficients in perceptual characteristics, pitch class profile in psychoacoustic characteristics and short-term energy in acoustic characteristics. In addition, we improve the neural networks classifier by fusing the LSTM unit into the convolutional neural networks. At last, we put the novel feature to the hybrid neural networks to recognize different vehicles. The results suggest the novel feature we proposed in this paper can increase the recognition rate by 7%; destroying the training data randomly by superimposing different kinds of noise can improve the anti-noise ability in our identification system; and LSTM has great advantages in modeling time series, adding LSTM to the networks can improve the recognition rate of 3.39%.
由于不同类型车辆模型的音频信息具有明显差异，因此可以通过车辆的音频信号准确识别车辆信息。在现实生活中，为了确定车辆的类型，我们不需要获取车辆的视觉信息，只需要获取音频信息。在本文中，我们从不同方面提取和拼接不同的特征：感知特征中的梅尔频率倒谱系数、心理声学特征中的音高类轮廓和声学特征中的短时能量。此外，我们通过将 LSTM 单元融合到卷积神经网络中来改进神经网络分类器。最后，我们将新特征应用于混合神经网络以识别不同的车辆。结果表明，本文提出的新特征可以将识别率提高 7%；通过叠加不同类型的噪声随机破坏训练数据，可以提高识别系统的抗噪声能力；LSTM 在时间序列建模方面具有很大的优势，在网络中添加 LSTM 可以将识别率提高 3.39%。
