A Novel Deep Learning Model for Indoor-Outdoor Scene Classification Using VGG-16 Deep CNN

Deepika Bhardwaj, Vinod Todwal

Scene Understanding, Scene Classification, Indoor – Outdoor Classification, Deep Learning, CNN, VGG-16.

Machines have begun to rule human beings as machines are performed nearly all the task what people are capable of doing in today’s world. The description of the scene is one word that gains significance in that machines imitate a human being's behavior. Scene classification can either be conducted on indoor or outdoor scenarios by means of different aspect feature extraction techniques. Indoor/Outdoor scenes' classification is found to be more demanding in these two categories. Scene classification of indoor-outdoor approaches has a poor accuracy problem. This research aims to enhance the accuracy by using the Convolution Neural Network Model in VGG-16. Indoor/Outdoor scenario classification. This paper proposes a new approach to VGG-16 to classify images into their classes. The algorithm results are tested using the SUN397- indoor-outdoor dataset. The experimental data reveals that the methodology proposed is superior to the existing technology for indoor-outdoor scene classification. From experimental results, we create that model shown the accuracy of 93.66 percent for indoor classes & 98.91 percent for outdoor classes. Effective tests show the validity of the proposed method.