video_inference
video_inference copied to clipboard
SubarnaTripathi
mean field based inference for dense semantic video segmentation with higher order clique potentials (Pn Potts model)
So if I want to run the code with my image, should I replace all the processed data related to the original image? And where can I find the algorithm...
How to understand the key code 'higher_order[curr_pix_index*M_+k] += (ho_param1*costdata - ho_param2*higher_order_prob);' in the function of "DenseCRF::calculate_higherorder_pot(int layer)" in the line 1270 of densecrf.cpp? And What do these four variables(ho_param1, ho_param2,...
Metadata
Owner
Metadata
mean field based inference for dense semantic video segmentation with higher order clique potentials (Pn Potts model)