Speaker Recognition Controversy
Prof Dr P Chandra Sekharan, President, Forensics International in a press release commented about the controversial CD said to have contained the recordings of the conversation between Shanti Bhushan and Mulayam singh thus: ‘The Central Forensic Science Laboratory seems to have opined that the CD is not doctored while the Hyderabad Truth Labs had said it is fabricated. But it is not clear whether both these agencies have examined CDs containing identical recordings i.e. both were true copies from the questioned CD said to contain the conversation between Shanti Bhushan and Mulayam Singh’.
There shall be controversy only if both the labs had examined the same contents and came out with contradicting reports. This important aspect has to be established before attributing authenticity or otherwise to the respective reports.
In as much as the recordings in a CD can only be a secondary recording either from the cell phones or sound recorders which will always contain the inherent defects due to copying. Sometimes the secondary recording may introduce volume changes; suppression of noise etc, besides wilful editing, addition, discontinuity, splicing, alteration and repetition. Therefore the finding that the CD is doctored or fabricated has no forensic evidentiary value.
What is important in this case is the speaker identification. Speaker recognition uses the acoustic features of speech that have been found to differ between individuals. These acoustic patterns reflect both anatomy , namely size and shape of the throat and mouth and learned behavioural patterns, namely voice pitch, speaking style. Speaker verification has earned speaker recognition its classification as a "behavioural biometric."
There are various technologies used to process and store voice prints. They include i) frequency estimation, ii) hidden Markov models, iii) Gaussian mixture models, iv) pattern matching algorithms, v) neural networks, vi) matrix representation, vii) Vector Quantization and viii) decision trees. Some systems also use "anti-speaker" techniques, such as cohort models, and world models.
Multi-Speech, Model 3700, GoldWave 5.06 and ALIZE/SpkDet softwares are the latest brand names which use ‘state- of- the- art’ open source speaker recognition.
Let both the CDs be sent to labs in France or Germany
