def __init__(self, obj_proposal_module=None, gpu_id=0, model='hand5', cpu_mode=False):
#Caffe Setting
self._model = model
self._gpu_id = gpu_id
self._cpu_mode = cpu_mode
#Objec Proposal Module
if obj_proposal_module is None:
self._obj_proposal_module = Selective_Search()
else:
self._obj_proposal_module = obj_proposal_module
#Config Caffe Model Path
self._prototxt = os.path.join(cfg.ROOT_DIR, 'models', NETS[self._model][0], 'test.prototxt')
self._caffemodel = os.path.join(cfg.ROOT_DIR, 'data', 'fast_rcnn_models',
NETS[self._model][1])
if not os.path.isfile(self._caffemodel):
raise IOError(('{:s} not found !!').formata(self._caffemodel))
#Config CPU or GPU Mode
if cpu_mode:
caffe.set_mode_cpu()
else:
caffe.set_mode_gpu()
caffe.set_device(self._gpu_id)
#Create Caffe
self._net = caffe.Net(self._prototxt, self._caffemodel, caffe.TEST)
print '\n\nLoaded network {:s}'.format(self._caffemodel)
class Fast_RCNN_Caffe(object):
def __init__(self, obj_proposal_module=None, gpu_id=0, model='hand5', cpu_mode=False):
#Caffe Setting
self._model = model
self._gpu_id = gpu_id
self._cpu_mode = cpu_mode
#Objec Proposal Module
if obj_proposal_module is None:
self._obj_proposal_module = Selective_Search()
else:
self._obj_proposal_module = obj_proposal_module
#Config Caffe Model Path
self._prototxt = os.path.join(cfg.ROOT_DIR, 'models', NETS[self._model][0], 'test.prototxt')
self._caffemodel = os.path.join(cfg.ROOT_DIR, 'data', 'fast_rcnn_models',
NETS[self._model][1])
if not os.path.isfile(self._caffemodel):
raise IOError(('{:s} not found !!').formata(self._caffemodel))
#Config CPU or GPU Mode
if cpu_mode:
caffe.set_mode_cpu()
else:
caffe.set_mode_gpu()
caffe.set_device(self._gpu_id)
#Create Caffe
self._net = caffe.Net(self._prototxt, self._caffemodel, caffe.TEST)
print '\n\nLoaded network {:s}'.format(self._caffemodel)
def detect_object(self, image):
#print 'fast_rcnn_caffe call'
return self.internal_detect_object(image, ('hand5',))
def internal_detect_object(self, image, classes):
# Load Object proposals
#timer = Timer()
#timer.tic()
obj_proposals = self._obj_proposal_module.get_object_proposals(image)
#timer.toc()
# Show Object proposals information
#print 'it took {:.3f}s for image ceate {:d} obj_proposals'.format(timer.total_time, obj_proposals.shape[0])
# Detect all object classes and regress object bounds
#timer = Timer()
#timer.tic()
scores, boxes = im_detect(self._net, image, obj_proposals)
#timer.toc()
#print ('Detection took {:.3f}s for '
# '{:d} object proposals').format(timer.total_time, boxes.shape[0])
# Visualize detections for each class
CONF_THRESH = 0.9
NMS_THRESH = 0.3
#Dictionary Result for Detection
class_detections = {}
#Extract detect result for each Class
for cls in classes:
cls_ind = CLASSES.index(cls)
cls_boxes = boxes[:, 4*cls_ind:4*(cls_ind + 1)]
cls_scores = scores[:, cls_ind]
keep = np.where(cls_scores >= CONF_THRESH)[0]
cls_boxes = cls_boxes[keep, :]
cls_scores = cls_scores[keep]
dets = np.hstack((cls_boxes,
cls_scores[:, np.newaxis])).astype(np.float32)
keep = nms(dets, NMS_THRESH)
dets = dets[keep, :]
#print 'All {} detections with p({} | box) >= {:.1f}'.format(cls, cls, CONF_THRESH)
#get current class detect result
detections = self.validate_detections(image, cls, dets, thresh=CONF_THRESH)
if len(detections) is not 0:
class_detections[cls] = detections
#Return Detection Result
return class_detections
def validate_detections(self, img, class_name, dets, thresh=0.5):
"""detected bounding boxes."""
customize_dets = dets[(dets[:, -1] >= thresh)]
#print 'Customize dets: {} len: {}'.format(customize_dets, len(customize_dets))
return customize_dets
"""
