Security in smart cities: A brief review of digital forensic schemes for biometric data

Ross, Arun; Banerjee, Sudipta; Chowdhury, Anurag

doi:10.1016/j.patrec.2020.07.009

Cited by 61 publications

(28 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In this work, we demonstrate DeepTalk's ability to reliably model the vocal style of a given speaker and transfer it to a synthetic speech with high fidelity. While this technique can improve the user-experience of Speech Generating Devices (SGD) [24] and digital voice assistants, several concerns are raised by its potential misuse for creating Deep-Fake speech. For example, in the past, DeepFake speech has been used to mimic an influential person's voice for defrauding [25].…”

Section: Ethical Implicationsmentioning

confidence: 99%

DEEPTALK: Vocal Style Encoding for Speaker Recognition and Speech Synthesis

Chowdhury

Ross

David

2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

Automatic speaker recognition algorithms typically characterize speech audio using short-term spectral features that encode the physiological and anatomical aspects of speech production. Such algorithms do not fully capitalize on speakerdependent characteristics present in behavioral speech features. In this work, we propose a prosody encoding network called DeepTalk for extracting vocal style features directly from raw audio data. The DeepTalk method outperforms several state-of-the-art speaker recognition systems across multiple challenging datasets. The speaker recognition performance is further improved by combining DeepTalk with a state-of-the-art physiological speech feature-based speaker recognition system. We also integrate DeepTalk into a current state-of-the-art speech synthesizer to generate synthetic speech. A detailed analysis of the synthetic speech shows that the DeepTalk captures F0 contours essential for vocal style modeling. Furthermore, DeepTalk-based synthetic speech is shown to be almost indistinguishable from real speech in the context of speaker recognition.

show abstract

Section: Ethical Implicationsmentioning

confidence: 99%

DEEPTALK: Vocal Style Encoding for Speaker Recognition and Speech Synthesis

Chowdhury

Ross

David

2021

ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Actually, most of the times, the work done by researchers on audio forensics addresses the problem of audio tampering. Ross, Banerjee, & Chowdhury (Ross et al, 2020) explained content-based audio manipulation techniques and the corresponding forensic detection methods, due to their direct impact on the biometric utility of the audio data that might affect the security in smart cities. They clarified how Digital audio manipulation can be deployed at multiple stages of audio production from audio recording such as tampering the sensor to audio editing where certain audio editing techniques such as butt splicing and copy-move forgery can be applied, thereby, compromising audio integrity.…”

Section: Literature Reviewmentioning

confidence: 99%

Verifying the Audio Evidence to Assist Forensic Investigation

Fayyad-Kazan¹,

Hejase²,

Moukadem³

et al. 2021

CIS

View full text Add to dashboard Cite

Audio forensics is a field in forensics that is used to authenticate, enhance, and analyze audio files to aid in solving different crime investigations. Audio as a forensic evidence must be enhanced and analyzed to be admissible in courts of law. But more importantly, it must be authenticated in order to prove that it is authentic and no manipulations were done to it. In this paper, an overview on audio forensics is presented, previous related work to this topic is shown, and methodologies for audio enhancement and authentication are explained along with audio tampering ways and signatures presentation.

show abstract

“…Metode penelitian dalam makalah ini yaitu dengan melakukan analisa komparasi antara suara barang bukti rekaman yang telah dimodifikasi oleh pelaku dengan suara pembanding asli [12], [13]. Skema penelitian digambarkan sebagai berikut: Berdasarkan data pada tabel 1 diketahui bahwa terdapat perbedaan nilai statistik pitch antara barang bukti dengan suara subjek walaupun tidak signifikan, artinya jika dilihat dari nilai minimum, maksimum, mean dan standar deviasi memiliki nilai statistik yang berbeda dan tidak identik.…”

Section: A Metode Forensikunclassified

Analisis Statisik Perbandingan Manipulasi Suara Dan Suara Asli Menggunakan Teknik Audio Forensik

Azzery

2020

teknokom

View full text Add to dashboard Cite

Kejahatan digital yang semakin beragam menuntut Tim ahli forensik untuk meningkatkan pengetahuan dalam pengungkapan kasus didalam dunia digital. Teknik audio forensik merupakan bagian dari ilmu digital forensik, yang lebih fokus pada analisa suara serta berbagai manipulasi didalamnya yang harus dapat dibuktikan di persidangan. Salah satu tantangan yang dihadapi dalam mengungkap kejahatan melalui audio yaitu adanya manipulasi suara yang berbeda dengan suara sumber atau pelaku. Analisa yang dilakukan menggunakan rekaman suara laki-laki yang terdiri dari 20 kata dan dilakukan manipulasi dengan menaikkan speed audio sebesar 20%. Metode yang digunakan yaitu dengan membandingkan hasil analisa Picth, Formant, dan Spectogram dari suara asli dan suara yang dimanipulasi. Hasil analisis perbandingan satistik Pitch, formant, spectogram menunjukkan bahwa terdapat perbedaan nilai dan range dari suara barang bukti dan suara subjek. Analisa statisik dilakukan dengan teknik One Way Annova menyatakan bahwa kedua suara rekaman tersebut Tidak Identik. Makalah ini diharapkan dapat menambah wawasan bagi Tim forensik untuk melakukan analisa lebih lanjut terhadap barang bukti yang sudah dimanipulasi.

show abstract

Security in smart cities: A brief review of digital forensic schemes for biometric data

Cited by 61 publications

References 29 publications

DEEPTALK: Vocal Style Encoding for Speaker Recognition and Speech Synthesis

DEEPTALK: Vocal Style Encoding for Speaker Recognition and Speech Synthesis

Verifying the Audio Evidence to Assist Forensic Investigation

Analisis Statisik Perbandingan Manipulasi Suara Dan Suara Asli Menggunakan Teknik Audio Forensik

Contact Info

Product

Resources

About