Authors : N. Poornima and B. Saleena
Abstract: One of the emerging technologies in applications like video recording and video compression holding significant importance over the years is video digitalization. Video retrieval is a popular research topic and various techniques are available in literature for the effective retrieval of videos. This research work presents a deep learning strategy based video retrieval scheme. Initially, the video archive is subjected for the key frame extraction, for extracting useful keyframes from the video. Then, the features have been extracted from the Keyframe and formulated as the feature database. The features are subjected for clustering using the Fuzzy C Means (FCM) algorithm. Then, clustered features have been provided to the deep learner for finding the optimal centroid for the incoming user query. For the experimentation, the research has considered videos from different category and both the text query and the video query have been used for the retrieval. Results from simulations demonstrate the efficiency of the proposed deep learning strategy in video retrieval and its achievement of improved values of 0.98 and 0.9743, respectively for recall, precision and F-measure.
N. Poornima and B. Saleena, 2020. An Automated Approach to Retrieve Lecture Videos using Context Based Semantic Features and Deep Learning. Journal of Engineering and Applied Sciences, 15: 3514-3525.