Cerebrum's purpose is getting continuous data inputs from different types of perceptions as memory sequences that triggered according to predefined threshold values and creating a complex time based relations between those memories by Crossmodal logic and training multiple Long Short-Term Memory Networks with this extracted data. Lastly creating outputs triggered by a stimuli, using pre-trained Artificial Neural Networks.