GPU-Based Integrated Security System for Minimizing Data Loss in Big Data Transmission

JPEG steganography with particle swarm optimization accelerated by AVX

“…Parallel and distributed computation have been successfully used to accelerate steganography and steganalysis . The acceleration aims at the computationally most expensive operations, including statistical image analysis, image feature extraction, classifier training, and error correction . It takes advantage of the fact that steganographic primitives such as convolution are, to a large extent, data parallel and suitable for an efficient execution on data‐parallel devices, including accelerators and graphic processing units (GPUs).…”

“…crucial for their practical applications. Parallel and distributed computation have been successfully used to accelerate steganography 7,8 and steganalysis. 6,9,10 The acceleration aims at the computationally most expensive operations, including statistical image analysis, 10 image feature extraction, 6,9,10 classifier training, 7,10 and error correction.…”

“…6,9,10 The acceleration aims at the computationally most expensive operations, including statistical image analysis, 10 image feature extraction, 6,9,10 classifier training, 7,10 and error correction. 8 It takes advantage of the fact that steganographic primitives such as convolution 9 are, to a large extent, data parallel and suitable for an efficient execution on data-parallel devices, including accelerators and graphic processing units (GPUs). Their high-performance parallel implementations have been shown to yield a speedup of one to three orders of magnitude.…”