The pretrained res10_300x300_ssd_iter_140000.caffemodel giving very good results for face detection. I want to implement that one from scratch in tensorflow. But not getting
