Authors : Dong-Hwan-Shin , Byung-Hoon Lee and Young-Seok Lee
Abstract: In recent years, interest in the broadcasting industry for second screen is increasing. This research focuses on ACR applications of audio watermarking at real world based on experimental results. In the ACR process, attenuation of a specific frequency band causes loss or distortion of the watermark embedded in the corresponding frequency band. Consequently, erroneous information is detected in the watermark detection. This study proposed an audio watermarking method based on experimental configuration related to the attenuation of the frequency band according to the distance obtained in the room where the TV speaker and the smartphone are installed. The proposed method is applied to 40 audio files in 5 sec to test imperceptibility and robustness. The experimental results show that the proposed method is robust to ADC and cropping attacks and imperceptible at human auditory system. The imperceptibility is proved by blind audience test at 92% and the watermark can be detectible from recorded audio signal above 90% at distance of 4 m and 60 db of signal intensity. And different segment length with the detection rate is compared. The longer the segment length, the greater the detection rate can be observed regardless types of music. The contribution of this research is that the proposed method was implemented on the experimental result of real world and simply computed while as high imperceptibility and robustness. Almost mobile and wearable devices in real world are required low computational complexity, low memory space and robustness in the analog world. Ultimately, the proposed method is suitable for the purposes of mobile and wearable devices in the real world. The experimental results show that the proposed method is robust to malicious attacks and imperceptible. The imperceptibility by blind audience test is 92% and detection rate is above 90%.
Dong-Hwan-Shin , Byung-Hoon Lee and Young-Seok Lee, 2017. Spread Spectrum Audio Watermarking Technology for Audio Content Recognition as Second Screen. Journal of Engineering and Applied Sciences, 12: 8127-8132.