부산대학교 도서관

Soft Computing is a set of techniques for addressing special class of complex problems which involve the use of computationally intelligent systems. The fusion of Deep Learning and Soft Computing tools like Neural Networks has shown a promising road ahead for computationally intensive problems, giving rise to algorithms and solutions with high efficiency and accuracy. Application of this fusion technique considerably improves the accuracy of recognizing faces in a facial recognition pipeline. Our hybrid model comprising of CNN & LSTM utilizes DCN'S which work on the concept of optimization of the embedding's of the face itself. Our proposed system has been tested on the CACD dataset and gives very promising & encouraging results