Video Captioning With Attention-Based LSTM and Semantic Consistency | IEEE Journals & Magazine | IEEE Xplore