University of Kalyani, India

 The usefulness of identifying a song from similar songs  is increasing with the growing  importance of high quality editing  and processing of audio signals. This paper proposes an effective  authentication technique for recognizing a song of a particular  singer from similar songs . Although a number of algorithms  have been developed for the identity and characteristics of a  song for recognizing artists, groups and musical genres, but the  authentication of  song for a particular singer has not yet been  fully utilized in computer audition algorithms. In this paper, we  present a framework for  identifying a particular song, which is,  sang by a particular singer by enforcing authentication techniques over the song without effecting the song quality. First  technique(Filtering of Low  Frequencies,ESKATS-FLF) is  developed based on the lower frequencies of the songs (which  are below the minimum frequency range of our electronic audio  player). In this technique we add extra value to the frequency  components which are below the minimum frequency to make  level of lower audible frequency. The second technique  (Embedding Secret Key, ESKATS-ESK ) is developed based on sampled data of song (PCM format) with embedding some  authenticate information over song without affecting song  quality of songs. In this technique we add some extra data in  some specific position of sampled data array and the specific  position is calculated by using sine wave characteristics. Experimental results are given based on Microsoft WAVE  (".wav") sound file.

1. [1] G.Erten, F.Salam,”Voice Output Extraction by Signal Separation ” ,  ISCAS ’98 ,ISBN 07803-4455-3,Vol 3, pp 5 - 8.

2. George Tzunetukis,” Song-Specific Bootstrapping of Singing Voice  Structure” 2004 IEEE International Conference on  Multimedia and Expo (ICME),Vol 5 ,pp 2027-2030.

3. Wei-Ho Tsai , Hsin-Min Wang,” A Query-by-example Framework to Retrieve Music  Documents by Singer” 2004 IEEE International Conference on Multimedia and Expo   ICME),Vol 5, pp 1863-1866.

4. George R. Doddington,” Speaker ecognition- Identifying People by  their Voices”, proceeding of the IEEE, VOL 73, NO.  11, NOV 1985,pp  1651-1665.

5. Qi Li,Biing –Hwabg Juang,Chin-Hui Lee,Qiru Zhou and Frank  K.Soong,”Recent Advancements in Automatic Spearker Authentication”, IEEE Robotics & Automation Magazine , March  1999,Vol 99,pp 24-34.

6. H. Suzuki, H. Zen, Y. Nunkuku, C.  Miyajima, K. Tokuda, and I.  Kitumuru,” Speech Recognition Using Voice-characteristic Dependent Acoustic Models”, ICASSP 2003,Vol 3, pp 740-743.

7. Kun S. Lin and Gene A. Frantz,” On Voice  Characteristics Conversion”, IEEE Transactions on Consumer Electronics, Vol. CE-30, No. 4,  November 1984, pp 598-603.