Multi-Stage Hiding of Image-into-Audio Steganography using CNN

Multi-Stage Hiding of Image-into-Audio Steganography using CNN

Mrs. T Madhavi Kumari ¹, Dr. A. Vinaya Babu², V. V. Chandra Teja³

¹(ECE, JNTUH College of Engineering Hyderabad, India)

²(Rtd. Prof. Of CSE, JNTU & Dean Academics, Stanley College of Engg. Tech, Hyderabad, India)

³(ECE, JNTUH College of Engineering Hyderabad, India)

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract – Sometimes a file consists of more information than it appears. A file may appear normal to an untrained eye, but expert recipients can extract more information from it. The increasing use of audio communication technology has accelerated the transfer of audio data via the Internet, making it a preferred carrier for covert communication. In this paper we propose a deep neural network based frame work to perform audio steganography. There are mainly two phases present, embedding phase and extraction phase. In the embedding phase the image contents are hidden in the audio carrier and in the extraction phase the hidden image is extracted from the embedded audio. The suggested framework will hide the secret message in a progressive manner, making the concealing process easier and capable of outperforming competing algorithms.

Key Words:  Audio steganography, Convolutional neural networks.