Explore the vast range of titles available.

This paper suggests a unique image encryption scheme based on key-based block ciphering followed by shuffling of ciphered bytes with variable-sized blocks, which makes this scheme substantially robust compared to other contemporary schemes available. Another distinguishing feature of this scheme is the usage of variable-sized key streams for consecutive blocks. Based on the elementary cellular automata with chaotic tent map, distinct key streams are used to cipher individual blocks. In the subsequent step, the bytes of the ciphered block so obtained are further shuffled to make the scheme more diffused. The block size varies with the varying key stream, which is again dependent on the preceding key stream as well as the plain image. It needs to be mentioned that the size of the first block and the key stream are generated from a 64-byte secret key and the plain image. Values of correlation and the number of pixel change rate between the original and the encrypted images are 0.000479 and 99.620901, respectively. Both of the above results along with other relevant experimental results strongly establish the robustness of the proposed scheme.

This paper presents an efficient image encryption scheme based on permutation followed by diffusion, where both of these phases use 2-d Sine logistic modulation map (SLMM) with different initial values. In addition, diffusion uses another map as 1-d Logistic chaotic map (LCM). The initial values of these chaotic maps are obtained from an external key of 64 bytes along with 32-byte hash value from the corresponding plain-image to incorporate plain-text sensitivity. Initially, confusion of the plain-image is implemented by applying row-level and column-level permutations. Then, this permuted image is used for subsequent diffusion, applied on block-level considering block size of 64 bytes. This diffusion process is accomplished by overlaying with chaotic matrix derived from LCM, followed by substitution of those overlaid bytes by DNA encoding along with SLMM to attain an encrypted image with an entropy nearly 8. Furthermore, all the chaotic values generated from the aforementioned maps are highly sensitive on the key as well as on the plain-image. This scheme is thoroughly verified on different sized plain-images with modern statistical analyses to prove the robustness of this scheme. Eventually, comparison with other schemes reinforces its competence and suitability to implement it in real-time system.

This paper presents a robust audio encryption scheme based on three consecutive phases, accomplished as cyclic shift followed by ciphering and wound up by shuffling, to break the high correlation amongst the neighbouring region of a plain audio. To encrypt a plain audio, the entire audio is split into different blocks of 64 bytes each and distinct key blocks are used for those audio blocks in the foregoing three phases. At first, the correlation of each audio block is reduced by the cyclic shift; thereafter, these shifted blocks are ciphered with piecewise linear chaotic map (PWLCM) along with elementary cellular automata (ECA) and finally, turns up with shuffling of ciphered bytes for better diffusion. The significant feature of this scheme is to generate distinct key blocks, which are highly sensitive to the secret key, a combination of the 64-byte external key along with the plain audio-dependent value. Moreover, these key blocks are derived using PWLCM from the secret key along with the preceding key block and previous encrypted block to achieve resistance against the known plain-text attack. It uses key space as large as 2 576 to resist brute-force attacks. The robustness as well as competence of this scheme is established with statistical analyses, cryptanalysis, randomness analysis and comparisons with existing schemes.

Increasing security concerns in crowd centric topologies have raised major interests in reliable face recognition systems globally. In this context, certain deep learning frameworks have been proposed till date, for example, Haar Cascade, MTCNN, Dlib to name a few. In this communication, we propose a deep neural network for reliable face recognition in high face density images. The proposed framework is inspired by multi-task cascaded convolutional neural Networks (MTCNN) and, hence the name MTCNN++. In this framework, we have modified the layer density with increasing the neuron count. All the three internal layers of MTCNN, viz. P-Net, R-Net, and O-Net layers and observe that the modified Net-Layer MTCNN (MTCNN++) perform equally well to the MTCNN library or better. Moreover, 20% dropout has been used for tuning the framework for better recognition of the faces, both in terms of face clarity and face count. MTCNN++ exhibits better results as the preprocessing is done dynamically in contrast to the previous versions. The training of the model was done on a dataset comprising of 113,586 human faces in a bucket of 9661 images. The comprehensive dataset comprised of photographs from varied events, thereby presenting multiple human expressions. The accuracy of the model varies from 87.7% (average of 12 faces per image) to 99.7% (average of 2 images per images). The proposed framework fares better with large face count per image. MTCNN++ has further been compared to other literary proposals, and the results are appreciable.

Increased use of internet demands substantial protection for secret image file from any adversary, specifically during transmission. In the field of cryptography there are two role models: cryptographer and crypt-analyst/attacker. The cryptographer develops techniques to make sure certain safety and security for transmissions while the crypt-analyst attempts to undo the former’s work by cracking the same. The basic goal of our scheme is to design an image encryption model which is extra challenging against any attack. In our research article, we have introduced session key dependent image encryption technique wherein the session key is the function of an original secret key (known for a pair of sender and receiver one time forever at the beginning) and the present secret image to be encrypted. Additionally the scheme does not require extracting and remembering of session keys to construct the subsequent session keys although the keys change during each transmission. Besides, in our scheme a double encryption technique is required, which once again confirms that the technique we propose is more robust than the conventional image encryption techniques known till date and is capable of resisting cyber-attacks of such kinds.

Multimedia file like audio demands special encryption technique due to its large data capacity without compromising correlation between it’s original and encrypted version (closer to zero). Most of the popular block cipher techniques work on multiple rounds whereas the proposed scheme guarantees the necessary low correlation between the original and the encrypted file without multiple rounds. The unique feature is that the consecutive blocks use different keys derived from the original one using the proposed key chaining algorithm and experimental results show that the correlation between the consecutive keys is also close to zero. The used encryption technique is based on DNA encoding with logistic chaotic map using the generated chain of keys. Furthermore, the concept of channel shuffling is introduced to make the encrypted data more secure. The experimental results confirm that the correlation between the original and ciphered block is close to zero and number of samples change rate value is close to 100. Again correlation between the two consecutive ciphered blocks is also close to zero, which conforms the acceptability of proposed scheme.

In this work, we formulated a real-world problem related to sewer pipeline gas detection using the classification-based approaches. The primary goal of this work was to identify the hazardousness of sewer pipeline to offer safe and non-hazardous access to sewer pipeline workers so that the human fatalities, which occurs due to the toxic exposure of sewer gas components, can be avoided. The dataset acquired through laboratory tests, experiments, and various literature sources was organized to design a predictive model that was able to identify/classify hazardous and non-hazardous situation of sewer pipeline. To design such prediction model, several classification algorithms were used and their performances were evaluated and compared, both empirically and statistically, over the collected dataset. In addition, the performances of several ensemble methods were analyzed to understand the extent of improvement offered by these methods. The result of this comprehensive study showed that the instance-based learning algorithm performed better than many other algorithms such as multilayer perceptron, radial basis function network, support vector machine, reduced pruning tree. Similarly, it was observed that multi-scheme ensemble approach enhanced the performance of base predictors.

Rapid growth of internet service attains better security of multimedia contents now a days. Heading this problem a DCT-based color image watermarking framework is proposed in this article. Many earlier works have suggested embedding watermark information in the low frequencies of the image to enhance the robustness against JPEG compression because low frequencies hold the most significant information of the image and not affected significantly by the quantization method of JPEG algorithm. Replacement of low-frequency components with watermark directly may incur undesirable degradation to the image quality. To preserve the visual quality of watermarked images, we are proposing a watermarking framework that adjusts the DCT low-frequency coefficients by scaled averaging. The security issue is well-taken care with double secret keys. Experimental result set demonstrates that the embedded watermark can be extracted efficiently from the JPEG-compressed images even after very high compression, re-watermarking, other image processing attacks. The extraction algorithm is blind i.e., neither host image nor the watermark is needed at the time of extraction.

This paper shows a symmetric image encryption based on bit-wise operation (XORing and Shifting). The basic idea is block ciphering (size of each block is 4 bytes) technique to cipher the secret bytes, after that ciphered bytes are again shuffled among N positions (N is the size of secret file). The scheme is combination of substitution as well as transposition techniques which provides additional protection of the secret data. The substitution and transposition are done using dynamic substitution box (SBOX) and transposition box (TBOX) which are generated using the secret key and made to vary for each block during ciphering. The size of encrypted data is same as the size of secret data and the proposed scheme has been tested using different images. We have also presented the security analysis such as key sensitivity analysis, statistical analysis, and differential analysis to prove the strength of our algorithm against crypto analysis.

In this work, we formulated a real-world problem related to sewer pipeline gas detection using the classification-based approaches. The primary goal of this work was to identify the hazardousness of sewer pipeline to offer safe and non-hazardous access to sewer pipeline workers so that the human fatalities, which occurs due to the toxic exposure of sewer gas components, can be avoided. The dataset acquired through laboratory tests, experiments, and various literature sources was organized to design a predictive model that was able to identify/classify hazardous and non-hazardous situation of sewer pipeline. To design such prediction model, several classification algorithms were used and their performances were evaluated and compared, both empirically and statistically, over the collected dataset. In addition, the performances of several ensemble methods were analyzed to understand the extent of improvement offered by these methods. The result of this comprehensive study showed that the instance-based learning algorithm performed better than many other algorithms such as multilayer perceptron, radial basis function network, support vector machine, reduced pruning tree. Similarly, it was observed that multi-scheme ensemble approach enhanced the performance of base predictors.