With increased human-machine interactions, computing systems are designed based on context and content for seamless user interactions. Nowadays, for efficient and personalized interactions, the user's emotional state plays a vital role in the design phase of the systems. Over the past decade, emotion recognition has garnered significant interest among researchers, with electroencephalogram (EEG) signals identified as a promising modality for distinguishing emotions. Among emotion recognition research, electroencephalogram (EEG) signals are found to be a potential modality to distinguish one's emotions. Several attempts have been made by researchers to increase the performance of the emotion recognition models. Hence, in this work, we proposed an EEG-based emotion recognition model that utilizes windowing techniques to classify emotions in the valence and arousal dimensions. We employed both non-overlapping and overlapping windowing techniques to segment the EEG signals. From these segmented EEG trials, a set of temporal and spatial features is extracted and used to train the classifier models. The proposed model achieves classification accuracies of 98.3 % and 98.2 % for valence and arousal dimensions, respectively. The experimental results and analysis reveal that the proposed model outperformed the state-of-the-art EEG-based emotion recognition studies on the AMIGOS dataset. Further, the proposed model can be employed in smart industries for effective human-machine interactions.

It is widely recognized that the public-key cryptosystems are playing tremendously an important role for providing the security services. In majority of the cryptosystems the crucial arithmetic operation is modular exponentiation. It is composed of a series of modular multiplications. Hence, the performance of any cryptosystem is strongly depends on the efficient implementation of these operations. This paper presents the Bit Forwarding 3-bits(BFW3) technique for efficient implementation of modular exponentiation. The modular multiplication involved in BFW3 is evaluated with the help of Montgomery method. These techniques improves the performance by reducing the frequency of modular multiplications. Results shows that the BFW3 technique is able to reduce the frequency of multiplications by 18.20% for 1024-bit exponent. This reduction resulted in increased throughput of 18.11% in comparison with MME42_C2 at the cost of 1.09% extra area. The power consumption reduced by 8.53% thereby saving the energy up to 10.10%.