Abstract This paper aims to contribute in visual computing over the implementation of face recognition algorithms on Jetson TK1. This powerful tool is a NVIDIA’s embedded Linux development platform featuring a Tegra K1 SOC which contains a CPU, a GPU and an ISP in a single chip. The Tegra K1 ‘class one’ is very faster than the former Tegra 2 ‘class two’ and... This paper aims to contribute in visual computing over the implementation of face recognition algorithms on Jetson TK1. This powerful tool is a NVIDIA’s embedded Linux development platform featuring a Tegra K1 SOC which contains a CPU, a GPU and an ISP in a single chip. The Tegra K1 ‘class one’ is very faster than the former Tegra 2 ‘class two’ and is the most efficient GPU in the ARM world. This particularity added to some optimizations we have made on implementation strategy and on recognition face algorithms conduct us to reach good results in real time systems RTS. Read More Read Less

Abstract Facial Expression Recognition is an application used for biometric software that can be used to recognize special expressions in a digital image by analysing and comparing different patterns. These software are popularly used for the purpose of security and are commonly used in other applications such as credit card verification, surveillance systems, medicines, home security, human-computer interface, etc.. Recognizing faces becomes very... Facial Expression Recognition is an application used for biometric software that can be used to recognize special expressions in a digital image by analysing and comparing different patterns. These software are popularly used for the purpose of security and are commonly used in other applications such as credit card verification, surveillance systems, medicines, home security, human-computer interface, etc.. Recognizing faces becomes very difficult when there is a frequent change occurs in facial expressions. In this paper two layer extension of HMM is used to identify continuous effective facial expressions. Partition technique is used for feature extraction. Two layered extension of HMM consists of bottom layer which represents the atomic expression made by eyes, nose and lips. Further upper layer represents the combination of these atomic expressions such as smile, fear etc. In HMM, Baum-Welch, Viterbi and Forward methods are used for parameter estimation for calculating the optimal state sequence and probability of the observed sequence respectively. This proposed system consists of three level of classification. Output of the first level is used for the training purposes for the second level and further this level is used for the third level for testing. Six basic facial expressions are recognised i.e. anger, disgust, fear, joy, sadness and surprise. Experimental result shows that Proposed System performs better than normal HMM and has the overall accuracy of 85% using JAFFE database. Read More Read Less

Abstract A analysis of human facial images has become increasingly important due to its numerous applications. In this regards, extracting facial parameter is vital and various studies have been done in this field. Hence in our proposed work, first time up to our knowledge, a robust automatic method is introduced for determining facial angles from profile view images using radon transform. Radon... A analysis of human facial images has become increasingly important due to its numerous applications. In this regards, extracting facial parameter is vital and various studies have been done in this field. Hence in our proposed work, first time up to our knowledge, a robust automatic method is introduced for determining facial angles from profile view images using radon transform. Radon transform is a kind of linear integration along a specific direction and angles play an important role to do this transform. The global features were rather considered by constructing a linear discriminant analysis (LDA) and also local features were rather considered by locality preserving projection (LPP). Our proposed combined algorithm has not only good precision, but also efficient performance and robust with noisy, scale and rotated image environments. In this work, several experiments have been conducted to analyze the robustness of our proposed Radon Combined Global and Local Preserving Features (RCGLPF) algorithm along with other existing conventional algorithms. Read More Read Less