Multi-style learning with denoising autoencoders for acoustic modeling in the internet of things (IoT)
Keywords: یادگیری نمایندگی; Deep learning; Deep neural networks; Multi-style training; Deep denoising autoencoders; Mixed training; Representation learning; Data combination; Data synthesis; Noise injection theory; Feature compensation; Automatic speech recognition; Internet of thin