A Long Sequence Speech Perceptual Hashing Authentication Algorithm Based on Constant Q Transform and Tensor Decomposition

Huang, Yibo; Hou, Hexiang; Zhang, Yuan; Fan, Manhong

doi:10.1109/access.2020.2974029

Cited by 12 publications

(3 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fig. 7 shows the comparison results of P-R curves between the proposed method and Li's method (2021) [25], Zhang's method (2023) [28], and Huang's method (2022) [39] methods.…”

Section: Speech Retrieval Performance Analysismentioning

confidence: 99%

Multi-User and multi-keyword verifiable encrypted speech retrieval method based on blockchain

Li,

Zhang,

Huang

et al. 2023

Preprint

View full text Add to dashboard Cite

Aiming at the existing encrypted speech retrieval schemes that do not support multi-user multi-keyword retrieval, low retrieval efficiency and accuracy, lack of verifiability of encrypted speech retrieval results, and user fairness, a multi-user and multi-keyword verifiable encrypted speech retrieval method based on blockchain was proposed. Firstly, use the advanced encryption standard-128 (AES-128) to encrypt the speech and upload it to the cloud server. Secondly, the Mel-frequency cepstral coefficient (MFCC) and Filter bank (Fbank) speech features are extracted and fused into new MFCC-Fbank features, which are input into the designed CNN-BiGRU model for training. These features are used to extract more robust deep features to construct hash codes and are used as searchable encrypted keywords. Finally, use smart contracts to store encrypted speech hash codes and corresponding index values and manage multi-user keys through trusted institutions. When retrieval, smart contracts are used as trusted third parties for retrieval work, and the designed adaptive homomorphic Message Authentication Code (MAC) algorithm is used to verify the correctness of the retrieval results, ensuring fairness in verification and data privacy. The experimental results show that the proposed method effectively prevents user privacy leakage and improves encrypted speech retrieval accuracy and data security under multi-user and multi-keyword conditions.

show abstract

“…Fig. 7 shows the comparison results of P-R curves between the proposed method and Li's method (2021) [25], Zhang's method (2023) [28], and Huang's method (2022) [39] methods.…”

Section: Speech Retrieval Performance Analysismentioning

confidence: 99%

Multi-User and multi-keyword verifiable encrypted speech retrieval method based on blockchain

Li,

Zhang,

Huang

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Table 9 shows the results for some selected state-of-the-art methods. [24] 0.0264 0.0366 0.781 0.758 [40] 0.0316 0.0341 0.637 0.944…”

Section: Comparison With Related Workmentioning

confidence: 99%

Robust Speech Hashing for Digital Audio Forensics

2019

View full text Add to dashboard Cite

The verification of the integrity and authenticity of multimedia content is an essential task in the forensic field, in order to make digital evidence admissible. The main objective is to establish whether the multimedia content has been manipulated with significant changes to its content, such as the removal of noise (e.g., a gunshot) that could clarify the facts of a crime. In this project we propose a method to generate a summary value for audio recordings, known as hash. Our method is robust, which means that if the audio has been modified slightly (without changing its significant content) with perceptual manipulations such as MPEG-4 AAC, the hash value of the new audio is very similar to that of the original audio; on the contrary, if the audio is altered and its content changes, for example with a low pass filter, the new hash value moves away from the original value. The method starts with the application of MFCC (Mel-frequency cepstrum coefficients) and the reduction of dimensions through the analysis of main components (principal component analysis, PCA). The reduced data is encrypted using as inputs two values from a particular binarization system using Collatz conjecture as the basis. Finally, a robust 96-bit code is obtained, which varies little when perceptual modifications are made to the signal such as compression or amplitude modification. According to experimental tests, the BER (bit error rate) between the hash value of the original audio recording and the manipulated audio recording is low for perceptual manipulations, i.e., 0% for FLAC and re-quantization, 1% in average for volume (−6 dB gain), less than 5% in average for MPEG-4 and resampling (using the FIR anti-aliasing filter); but more than 25% for non-perceptual manipulations such as low pass filtering (3 kHz, fifth order), additive noise, cutting and copy-move.

show abstract

“…literature 3 proposed a frequency band variancebased speech-aware hash retrieval method with integrity and security in decrypting ciphertext data but poor template diversity and revocability. The literature 4 proposes long sequence speech-aware hashing methods with better robustness but long hash sequences increase the running time and retrieval efficiency. It can be seen that nowadays speech-aware hashing can no longer meet the demand of speech content authentication technology, and BioHash is a unique perceptual hash function with higher security than perceptual hashing and lower spatial complexity than quantum hashing.…”

Section: Introductionmentioning

confidence: 99%

Blockchain and IPFS-based authentication method for speech BioHash content

Gao,

Huang,

Hua

2023

International Conference on Image, Signal Processing, and Pattern Recognition (ISPP 2023)

View full text Add to dashboard Cite

Nowadays, there are various ways to protect the authenticity and reliability of voice data during transmission such as digital signature and voice encryption, but these ways inevitably change a small amount of voice frames and data content. In order to prevent problems such as voice data content being easily tampered and stolen during transmission and improve the security of voice data content in the cloud. In this paper, a voice BioHash content authentication method based on blockchain and IPFS is proposed by using the voice BioHash method and combining the blockchain technology and the characteristics of interstellar file system. The method first uses improved spectral entropy method and spectral flux for feature fusion, and Chebyshev-Sine function constitutes a pseudo-random matrix projection mapping and binarization processing to generate voice BioHash, store the BioHash in the block transaction information, upload the voice to the Interstellar File System, query the block transaction information to retrieve the voice related information during authentication, and return the authentication result by matching completion of authentication. The experimental results show that the method has high collision resistance and authentication efficiency, and the voice BioHash method has better differentiation and robustness.

show abstract

A Long Sequence Speech Perceptual Hashing Authentication Algorithm Based on Constant Q Transform and Tensor Decomposition

Cited by 12 publications

References 33 publications

Multi-User and multi-keyword verifiable encrypted speech retrieval method based on blockchain

Multi-User and multi-keyword verifiable encrypted speech retrieval method based on blockchain

Robust Speech Hashing for Digital Audio Forensics

Blockchain and IPFS-based authentication method for speech BioHash content

Contact Info

Product

Resources

About