* refine test reader return

* add infer and parse_args

return im_id in reader for evaluation 
add coco evaluation for Faster Rcnn model
visualize detection result when image_path is provided