• The Journal of the Acoustical Society of Korea
• /
• v.31 no.4
• /
• pp.260-265
• /
• 2012

A robust audio fingerprinting method is presented based on segmentation boundaries. In order to obtain robustness against linear speed changes, fingerprint extraction and matching are synchronized with the segmentation boundaries. Experimental results show that the proposed method is also robust against other common audio processing steps including low bit-rate compression, equalization, and time-scale modification.

• ETRI Journal
• /
• v.25 no.1
• /
• pp.52-54
• /
• 2003

We propose an improved algorithm for detecting audio watermarks based on a spread spectrum in the spectral domain. Since the energy of a watermark is much smaller than that of the cover audio data, pre-processing to reduce the effect of the cover data is needed to reliably extract watermarks. We introduce a spectral envelope filter as a pre-process that enhances detecting performance by filtering out the intrinsic spectral character of cover data. The proposed watermarking structure can be easily included in the compression system and can extract watermarks from partially decompressed spectral data. Our experimental results demonstrate that with a bit error rate of around 10 dB against general attacks, the proposed detecting scheme works better than detectors without the spectral filter.

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.8
• /
• pp.7-13
• /
• 2015

Recently, there are serviced user customized advertisement of various type using smart device. Representative services are advertisement service using light of smart TV screen or audible sound of smart TV to transmit advertisement information. However, those services have to do a specific action of smart device user for advertisement information or need audible audio information of TV contents. To overcome those weakness, therefore, we propose an advertisement method using inaudible sound of speaker based on smart device. This method supports the transfer of advertising content to the smart device user with no additional action or TV audio signal required to access that content. The proposed method used two high frequencies among 18kHz ~ 22kHz of audible frequency range which smart TV can send out. And it generates those frequencies synthesized with audio of TV contents as trigger signal which can send advertisements to smart device. Next, smart device analysis the trigger signal and request advertisement contents related to the signal to server. After then, smart device can show the downloaded contents to user. Because the proposed method uses the high frequencies of sound signals via the inner speaker of the smart device, its main advantage is that it does not affect the audio signal of TV content. To evaluate the efficacy of the proposed method, we developed an application to implement it and subsequently carried out an advertisement transmission experiment. The success rate of the transmission experiment was approximately 97%. Based on this result, we believe the proposed method will be a useful technique in introducing a customized user advertising service.

• The Journal of the Acoustical Society of Korea
• /
• v.33 no.3
• /
• pp.210-215
• /
• 2014

In spite of distortion caused by noise and echo, the audio fingerprinting technique must identify successfully an audio source. This audio fingerprinting technique is applying for TV commercial advertisement identification. In this paper, we propose a robust audio fingerprinting method for TV commercial advertisement identification. In the proposed method, a prominent audio peak pair fingerprint based on constant Q transform improves the accuracy of the audio fingerprinting system in real noisy environments. Experimental results confirm that the proposed method is quite robust than previous audio fingerprinting method in different noise conditions and achieves promising accurate results.

• The KIPS Transactions:PartB
• /
• v.19B no.1
• /
• pp.19-26
• /
• 2012

The rapid increase of information imposes new demands of content management. The purpose of automatic audio segmentation and classification is to meet the rising need for efficient content management. With this reason, this paper proposes a high-accuracy algorithm that segments audio signals and classifies them into different classes such as speech, music, silence, and environment sounds. The proposed algorithm utilizes support vector machine (SVM) to detect audio-cuts, which are boundaries between different kinds of sounds using the parameter sequence. We then extract feature vectors that are composed of statistical data and they are used as an input of fuzzy c-means (FCM) classifier to partition audio-segments into different classes. To evaluate segmentation and classification performance of the proposed SVM-FCM based algorithm, we consider precision and recall rates for segmentation and classification accuracy for classification. Furthermore, we compare the proposed algorithm with other methods including binary and FCM classifiers in terms of segmentation performance. Experimental results show that the proposed algorithm outperforms other methods in both precision and recall rates.

• Journal of Korea Society of Industrial Information Systems
• /
• v.12 no.1
• /
• pp.54-60
• /
• 2007

The Digital Multimedia Broadcasting(DMB) system is developed to offer high quality multimedia content to the mobile environment. The system adopts the MPEG-4 standard for the main video, audio and other media format. For providing interactive contents, it also adopts the MPEG-4 scene description that refers to the spatio-temporal specifications and behaviors of individual objects. With more interactive contents, the scene description also needs higher bitrate. However, the bandwidth for allocating meta data, such as scene description is restrictive in the mobile environment. On one hand, the DMB terminal renders each media stream according to the scene description. Thus the binary format for scene(BIFS) stream corresponding to the scene description should be decoded and parsed in advance when presenting media data. With this reasoning, the transmission delay of the BIFS stream would cause the delay in transmitting whole audio-visual scene presentations, although the audio or video streams are encoded in very low bitrate. This paper presents the effective optimization technique in adapting the BIFS stream into the expected bitrate without any waste in bandwidth and avoiding transmission delays inthe initial scene description for interactive DMB content.

• ETRI Journal
• /
• v.35 no.6
• /
• pp.960-968
• /
• 2013

To implement a cloud game service platform supporting multiple users and devices based on real-time streaming, there are many technical needs, including game screen and sound capturing, audio/video encoding in real time created by a high-performance server-generated game screen, and real-time streaming to client devices, such as low-cost PCs, smart devices, and set-top boxes. We therefore present a game service platform for the running and management of the game screen, as well as running the sound on the server, in which the captured and encoded game screen and sound separately provide client devices through real-time streaming. The proposed platform offers Web-based services that allow game play on smaller end devices without requiring the games to be installed locally.

• Journal of Korea Society of Industrial Information Systems
• /
• v.11 no.3
• /
• pp.10-14
• /
• 2006

To effectively manage a very large amount of music data, we need an indexing technique based on audio features. But the indexing techniques for audiofeatures have not been studied completely. In this paper, we describe a content-based music information retrieval technique for audio features using the TIP-indexing file. In addition, we develop and experiment the TIP-indexing files using various blocking factors to present performance comparisons for effective indexing. Experimental results show the effectiveness of the proposed techniques.

• Proceedings of the IEEK Conference
• /
• 2004.08c
• /
• pp.768-771
• /
• 2004

In this paper, we propose a content-based audio genre classification algorithm that automatically classifies the query audio into five genres such as Classic, Hiphop, Jazz, Rock, Speech using digital signal processing approach. From the 20 second query audio file, 54 dimensional feature vectors, including Spectral Centroid, Rolloff, Flux, LPC, MFCC, is extracted from each query audio. For the classification algorithm, k-NN, Gaussian, GMM classifier is used. In order to choose optimum features from the 54 dimension feature vectors, SFS (Sequential Forward Selection) method is applied to draw 10 dimension optimum features and these are used for the genre classification algorithm. From the experimental result, we verify the superior performance of the SFS method that provides near $90{\%}$ success rate for the genre classification which means $10{\%}$-$20{\%}$ improvements over the previous methods

• 한국정보통신설비학회:학술대회논문집
• /
• 2006.08a
• /
• pp.219-222
• /
• 2006

In this paper we present a cryptanalysis of the generalized patchwork algorithm under the assumption that the attacker possesses only a single copy of the watermarked audio. In the scheme, watermark is inserted by modifying randomly chosen DCT values in each block of the original audio. Towards the attack we first fit low degree polynomials (which minimize the mean square error) on the data available from each block of the watermarked content. Then we replace the corresponding DCT data of the at-tacked audio by the available data from the polynomials to construct an attacked audio. The technique nullifies the modification achieved during watermark embedding. Experimental results show that recovery of the watermark becomes difficult after the attack.