administrator/frigate
1
0
Fork 0
mirror of https://github.com/blakeblackshear/frigate.git synced 2026-02-03 01:35:22 +03:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

Load model and submit for inference.

Browse Source 
Sucessfully load model and initialize OpenVino engine with either CPU or GPU as device.
Does not parse results for objects.
This commit is contained in:
Nate Meyer 2022-08-26 02:00:11 -04:00
parent 4b675692ec
commit cf03f62088
3 changed files with 41 additions and 69 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
docker/install_deps.sh
View File

@ -53,6 +53,7 @@ if [[ "${TARGETARCH}" == "amd64" ]]; then
echo 'deb http://deb.debian.org/debian testing main non-free' >/etc/apt/sources.list.d/debian-testing.list
apt-get -qq update
apt-get -qq install --no-install-recommends --no-install-suggests -y \
intel-opencl-icd \
mesa-va-drivers libva-drm2 intel-media-va-driver-non-free i965-va-driver libmfx1 radeontop intel-gpu-tools
rm -f /etc/apt/sources.list.d/debian-testing.list
fi

105
frigate/detectors/openvino.py
View File

@ -10,79 +10,48 @@ logger = logging.getLogger(__name__)
class OvDetector(DetectionApi):
def __init__(self, det_device=None, model_path=None, num_threads=1):
self.ovCore = ov.Core()
self.ovModel = self.ovCore.read_model(model_path)
self.interpreter = self.ovCore.compile_model(
model=self.ovModel, device_name=det_device
self.ov_core = ov.Core()
self.ov_model = self.ov_core.read_model(model_path)
self.interpreter = self.ov_core.compile_model(
model=self.ov_model, device_name=det_device
)
self.interpreter.allocate_tensors()
self.tensor_input_details = self.interpreter.get_input_details()
self.tensor_output_details = self.interpreter.get_output_details()
logger.info(f"Model Input Shape: {self.interpreter.input().shape}")
logger.info(f"Model Output Shape: {self.interpreter.output().shape}")
def detect_raw(self, tensor_input):
self.interpreter.set_tensor(self.tensor_input_details[0]["index"], tensor_input)
self.interpreter.invoke()
tensor_transpose = np.reshape(tensor_input, self.interpreter.input().shape)
boxes = self.interpreter.tensor(self.tensor_output_details[0]["index"])()[0]
class_ids = self.interpreter.tensor(self.tensor_output_details[1]["index"])()[0]
scores = self.interpreter.tensor(self.tensor_output_details[2]["index"])()[0]
count = int(
self.interpreter.tensor(self.tensor_output_details[3]["index"])()[0]
)
infer_request = self.interpreter.create_infer_request()
results = infer_request.infer([tensor_transpose])
# class_ids = self.interpreter.tensor(self.tensor_output_details[1]["index"])()[0]
# scores = self.interpreter.tensor(self.tensor_output_details[2]["index"])()[0]
# count = int(
# self.interpreter.tensor(self.tensor_output_details[3]["index"])()[0]
# # )
# class_ids = results[0, 0, :, 1]
# # class_ids = [0, 0, 1, 1]
# print(class_ids)
# scores = results
# print(scores)
detections = np.zeros((20, 6), np.float32)
for i in range(count):
if scores[i] < 0.4 or i == 20:
break
detections[i] = [
class_ids[i],
float(scores[i]),
boxes[i][0],
boxes[i][1],
boxes[i][2],
boxes[i][3],
]
return detections
class GpuOpenVino(DetectionApi):
def __init__(self, det_device=None, model_path=None, num_threads=1):
self.interpreter = tflite.Interpreter(
model_path=model_path or "/cpu_model.tflite", num_threads=3
)
self.interpreter.allocate_tensors()
self.tensor_input_details = self.interpreter.get_input_details()
self.tensor_output_details = self.interpreter.get_output_details()
def detect_raw(self, tensor_input):
self.interpreter.set_tensor(self.tensor_input_details[0]["index"], tensor_input)
self.interpreter.invoke()
boxes = self.interpreter.tensor(self.tensor_output_details[0]["index"])()[0]
class_ids = self.interpreter.tensor(self.tensor_output_details[1]["index"])()[0]
scores = self.interpreter.tensor(self.tensor_output_details[2]["index"])()[0]
count = int(
self.interpreter.tensor(self.tensor_output_details[3]["index"])()[0]
)
detections = np.zeros((20, 6), np.float32)
for i in range(count):
if scores[i] < 0.4 or i == 20:
break
detections[i] = [
class_ids[i],
float(scores[i]),
boxes[i][0],
boxes[i][1],
boxes[i][2],
boxes[i][3],
]
# i = 0
# for object_detected in results["detection_out"][0, 0, :]:
# if object_detected[2] < 0.1 or i == 20:
# break
# detections.append(
# [
# object_detected[1],
# float(object_detected[2]),
# object_detected[3],
# object_detected[4],
# object_detected[5],
# object_detected[6],
# ]
# )
# i += 1
return detections

4
requirements-wheels.txt
View File

@ -3,7 +3,8 @@ Flask == 2.2.*
imutils == 0.5.*
matplotlib == 3.6.*
mypy == 0.942
numpy == 1.22.*
# numpy == 1.22.*
numpy == 1.19.*
opencv-python-headless == 4.5.5.*
paho-mqtt == 1.6.*
peewee == 3.15.*
@ -18,3 +19,4 @@ scipy == 1.8.*
setproctitle == 1.2.*
ws4py == 0.5.*
zeroconf == 0.39.4
openvino == 2022.*