DWT-GBT-SVD-based Robust Speech Steganography
This work addresses network security for communications by providing a robust steganography method, though it appears incremental as it combines existing transforms.
The paper tackled the problem of secure communication by developing a speech steganography method that hides secret messages in audio signals, achieving imperceptibility and robustness against various attacks like noise and compression as demonstrated on the NOIZEUS database.
Steganography is a method that can improve network security and make communications safer. In this method, a secret message is hidden in content like audio signals that should not be perceptible by listening to the audio or seeing the signal waves. Also, it should be robust against different common attacks such as noise and compression. In this paper, we propose a new speech steganography method based on a combination of Discrete Wavelet Transform, Graph-based Transform, and Singular Value Decomposition (SVD). In this method, we first find voiced frames based on energy and zero-crossing counts of the frames and then embed a binary message into voiced frames. Experimental results on the NOIZEUS database show that the proposed method is imperceptible and also robust against Gaussian noise, re-sampling, re-quantization, high pass filter, and low pass filter. Also, it is robust against MP3 compression and scaling for watermarking applications.