Fig. 4From: Semantic context driven language descriptions of videos using deep neural networkPlot of training loss versus number of epochs.: a 2-layered LSTM b 3-layered LSTMBack to article page