08_june_2017

[@gitabcworld] Neural Message Passing: How the hidden states h^t_v are represented with respect to a CNN? In the case of the Devenaud approach, it is possible to draw a CNN similarity showing the W_t matrix and h^t_v layers?

[@priba] YOLO:Si les coordenades es prediuen al centre de la regio, com es detecten objectes que tinguin el mateix centre pero diferent mida?