问题
I train the CaffeNet (more precisely Cifar10 model for two classes classification) model. Now the model is ready for detection. For the model testing with a single image, I use test_predict_imagenet.cpp. I haven't tested how fast the code can run for 640 x 480 image. My target is I like to have 5~10 frame/sec is just nice for offline detection. I understand that I need to implement multi-size detection (i.e. something like we do in face detection, original image size is re-sized for different smaller sizes) so that I don't miss the pedestrian in each frame.
According to this paper, they use 64 x 128 image size in training and detection takes 3ms/window and for 100 windows/image, it takes 300msec/frame. Not sure they implement multi-size detection approach. If multi-size is implemented, it will take much longer.
At this moment, I have only knowledge of implementing test_predict_imagenet.cpp method for multi-size detection. I know it will be very slow.
Is there any more efficient way of detection using CaffeNet model? My target is just nice for 5~10 frame/sec rate. Thanks
来源：https://stackoverflow.com/questions/31992971/detecting-pedestrian-using-caffenet-model-at-moderate-framerate