if back_detector:
face_detector.load_weights("blazefaceback.pth")
face_detector.load_anchors("anchors_face_back.npy")
else:
face_detector.load_weights("blazeface.pth")
face_detector.load_anchors("anchors_face.npy")
palm_detector = BlazePalm().to(gpu)
palm_detector.load_weights("blazepalm.pth")
palm_detector.load_anchors("anchors_palm.npy")
palm_detector.min_score_thresh = .75
hand_regressor = BlazeHandLandmark().to(gpu)
hand_regressor.load_weights("blazehand_landmark.pth")
face_regressor = BlazeFaceLandmark().to(gpu)
face_regressor.load_weights("blazeface_landmark.pth")
WINDOW='test'
cv2.namedWindow(WINDOW)
if len(sys.argv) > 1:
capture = cv2.VideoCapture(sys.argv[1])
mirror_img = False
else:
capture = cv2.VideoCapture(0)
mirror_img = True
if capture.isOpened():
hasFrame, frame = capture.read()
frame_ct = 0
if back_detector:
face_detector.load_weights("blazefaceback.pth")
face_detector.load_anchors("anchors_face_back.npy")
else:
face_detector.load_weights("blazeface.pth")
face_detector.load_anchors("anchors_face.npy")
palm_detector = BlazePalm().to(gpu)
palm_detector.load_weights("blazepalm.pth")
palm_detector.load_anchors("anchors_palm.npy")
palm_detector.min_score_thresh = .75
hand_regressor = BlazeHandLandmark().to(gpu)
hand_regressor.load_weights("blazehand_landmark.pth")
face_regressor = BlazeFaceLandmark().to(gpu)
face_regressor.load_weights("blazeface_landmark.pth")
WINDOW = 'test'
cv2.namedWindow(WINDOW)
if len(sys.argv) > 1:
capture = cv2.VideoCapture(sys.argv[1])
mirror_img = False
else:
capture = cv2.VideoCapture(2)
mirror_img = True
if capture.isOpened():
hasFrame, frame = capture.read()
frame_ct = 0
else:
import torch
import cv2
import sys
from blazebase import resize_pad, denormalize_detections
from blazeface import BlazeFace
from blazepalm import BlazePalm
from blazeface_landmark import BlazeFaceLandmark
from blazehand_landmark import BlazeHandLandmark
from visualization import draw_detections, draw_landmarks, draw_roi, HAND_CONNECTIONS, FACE_CONNECTIONS
gpu = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
torch.set_grad_enabled(False)
model = BlazeFaceLandmark().to(gpu)
model.load_weights("blazeface_landmark.pth")
##############################################################################
batch_size = 1
height = 192
width = 192
x = torch.randn((batch_size, height, width, 3),
requires_grad=True).byte().to(gpu)
opset = 12
##############################################################################
input_names = ["input"] #[B,192,192,3],
output_names = ['landmark', 'confidence'] #[B,486,3], [B]
onnx_file_name = "BlazeFace_{}x{}x{}xBGRxByte_opset{}.onnx".format(