Fig. 2From: Semantic context driven language descriptions of videos using deep neural networkLSTM unit [40]Back to article page