Aim: The objective of this research is to observe the non-rapid eye movement expressions that have been detected using YOLO algorithm and compared with the Convolutional Neural Network. Materials and Methods: A YOLO algorithm is used to detect eye movement. YOLO has the special features which helps in classifying the class like circles, squares, colors and distance of the images. Two groups, containing YOLO and CNN were used to find the accuracy of novel non rapid eye movement detection with 15 samples each to calculate, a total of 30 samples were taken in this work. Sample size calculated using G power with pretest power at 80%, error rate of 0.05. Thus these features will help in finding the expression of NREM (Novel Non Rapid eye movement detection). Results: From the result obtained YOLO has accuracy 96.12% and CNN has accuracy 95.65%. YOLO has significantly high accuracy 0.013 (p<0.05) Conclusion: The detection rate is improved in terms of accuracy in YOLO which gives greater mean accuracy about 95.14% which is significantly better when compared with CNN in facial images.