CNN-LSTM architectures are a new RNN layer where the input of input transformations and recurrent transformations are both convolutional. Despite the very similar name, CNN-LSTM layers are therefore different from the combination of CNN and LSTM, as described above. The model is described in the paper Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting, Xingjian Shi, Zhourong Chen, Hao Wang, Dit-Yan Yeung, Wai-kin Wong, Wang-chun Woo, 2015,(https://arxiv.org/abs/1506.04214) and in 2017 some people are starting to experiment using this module for video, but this is still an area of active research.