@"jakaskerl"#1112 @"filipjeretina"#2315 , I am just curious to know if the following code can be used without service but still perform the same. Can you let me know if this is a possibility with the following code? Link to Code https://github.com/luxonis/depthai-experiments/tree/master/gen2-people-tracker Specific Code https://github.com/luxonis/depthai-experiments/blob/master/gen2-people-tracker/main.py From Michael

Hi @"michaelsalerno"#2322 , I'm not sure I understand what you mean by "without service", can you elaborate? Thanks!, Filip

Dear @"filipjeretina"#2315 , Yes, if my laptop did not have any wifi, but I still wanted to use the program the same. If I was in a remote area without wifi or a way to connect my laptop to the internet. From Michael Salerno

@"filipjeretina"#2315 is this a possibility?

@"jakaskerl"#1112 What is it using for the segmentation? A yolo model?

No service use

michaelsalerno

@filipjeretina

is this a possibility?

michaelsalerno

@filipjeretina

I know when you load it up it says something about converting the blobs via a website? But I want to know if it is possible to have it run offline, no connection to the internet, to maintain consistency. I do not want to have the image recognition be based on the connection, but rather its ability to recognize me.

jakaskerl

michaelsalerno

michaelsalerno I do not want to have the image recognition be based on the connection, but rather its ability to recognize me.

This is not the case even if you have network connection. The blobs (models) are trained to recognize human shapes. They are static and don't change. If the blob you need in the script is downloaded, you can use it without internet connection.

Or am I misunderstanding?

Thanks,
Jaka

michaelsalerno

Dear @jakaskerl ,

I have tried it on my end, and i get an error stating that it can't connect to the API.

From Michael

jakaskerl

michaelsalerno
Can you paste this error you are getting?

michaelsalerno What is it using for the segmentation? A yolo model?

The example you linked to is not using segmentation, only detection using a MBnet model. https://docs.openvino.ai/2024/omz_models_model_person_detection_retail_0013.html

Thanks,
Jaka

michaelsalerno

@jakaskerl

What is it using for the segmentation? A yolo model?

michaelsalerno

@jakaskerl

I want to see what the code is using, and I want to see what images and know how many images this is trained on.

michaelsalerno

Traceback (most recent call last):

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\urllib3\connection.py", line 174, in _new_conn

conn = connection.create_connection(

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\urllib3\util\connection.py", line 72, in create_connection

for res in socket.getaddrinfo(host, port, family, socket.SOCK_STREAM):

File "C:\Program Files\WindowsApps\PythonSoftwareFoundation.Python.3.10_3.10.3056.0_x64__qbz5n2kfra8p0\lib\socket.py", line 955, in getaddrinfo

for res in _socket.getaddrinfo(host, port, family, type, proto, flags):

socket.gaierror: [Errno 11001] getaddrinfo failed

During handling of the above exception, another exception occurred:

Traceback (most recent call last):

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\urllib3\connectionpool.py", line 715, in urlopen

httplib_response = self._make_request(

self._validate_conn(conn)

conn.connect()

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\urllib3\connection.py", line 363, in connect

self.sock = conn = self._new_conn()

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\urllib3\connection.py", line 186, in _new_conn

raise NewConnectionError(

urllib3.exceptions.NewConnectionError: <urllib3.connection.HTTPSConnection object at 0x000002065E4564D0>: Failed to establish a new connection: [Errno 11001] getaddrinfo failed

During handling of the above exception, another exception occurred:

Traceback (most recent call last):

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\requests\adapters.py", line 486, in send

resp = conn.urlopen(

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\urllib3\connectionpool.py", line 801, in urlopen

retries = retries.increment(

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\urllib3\util\retry.py", line 594, in increment

raise MaxRetryError(_pool, url, error or ResponseError(cause))

urllib3.exceptions.MaxRetryError: HTTPSConnectionPool(host='blobconverter.luxonis.com', port=443): Max retries exceeded with url: /zoo_models?version=2022.1&zoo_type=intel (Caused by NewConnectionError('<urllib3.connection.HTTPSConnection object at 0x000002065E4564D0>: Failed to establish a new connection: [Errno 11001] getaddrinfo failed'))

During handling of the above exception, another exception occurred:

Traceback (most recent call last):

File "C:\Users\sale4088\AppData\Local\Programs\Depthai2\depthai-viewer\scripts\depthai-experiments\gen2-people-tracker\main.py", line 12, in <module>

tracker = oak.create_nn('person-detection-retail-0013', color_cam, tracker=True)

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\depthai_sdk\oak_camera.py", line 146, in create_nn

comp = NNComponent(self._pipeline,

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\depthai_sdk\components\nn_component.py", line 100, in init

self._parse_model(model)

zoo_models = blobconverter.zoo_list()

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\blobconverter\init.py", line 505, in zoo_list

response = requests.get("{}/zoo_models?{}".format(url, urllib.parse.urlencode(url_params)))

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\requests\api.py", line 73, in get

return request("get", url, params=params, \*\*kwargs)

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\requests\api.py", line 59, in request

return session.request(method=method, url=url, \*\*kwargs)

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\requests\sessions.py", line 589, in request

resp = self.send(prep, \*\*send_kwargs)

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\requests\sessions.py", line 703, in send

r = adapter.send(request, \*\*kwargs)

File "C:\Users\sale4088\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\LocalCache\local-packages\Python310\site-packages\requests\adapters.py", line 519, in send

raise ConnectionError(e, request=request)

requests.exceptions.ConnectionError: HTTPSConnectionPool(host='blobconverter.luxonis.com', port=443): Max retries exceeded with url: /zoo_models?version=2022.1&zoo_type=intel (Caused by NewConnectionError('<urllib3.connection.HTTPSConnection object at 0x000002065E4564D0>: Failed to establish a new connection: [Errno 11001] getaddrinfo failed'))

michaelsalerno

@jakaskerl

How many images is that model trained off of?

jakaskerl

michaelsalerno
It's an SDK issue, yes.

Add:

models = getSupportedModels(printModels=False)
            try:
                zoo_models = blobconverter.zoo_list()
            except requests.exceptions.ConnectionError:
                zoo_models = []

to depthai_sdk\components\nn_component.py line 230

michaelsalerno How many images is that model trained off of?

I don't know, there isn't any info on the website.

Thanks,
Jaka

michaelsalerno

@jakaskerl ,

When connecting to the internet, and using the OPENVINO model, is this technically cloud computing?

jakaskerl

michaelsalerno
No, cloud computing means you are running inference on some other machine in the cloud. In the case of depthai, the model is loaded on the OAK device and is running on RVC VPU. This is called edge computing.

Thanks,
Jaka

michaelsalerno

@jakaskerl

Thank you, I really appreciate it!

michaelsalerno

@jakaskerl

If it is still only working through wifi, does that mean its edge computing still? Or does this mean that it is cloud computing?

jakaskerl

michaelsalerno
Just because the OAK device is connected via WiFi does not mean it’s cloud computing. WiFi is simply the medium for network connectivity. If the inference is still happening locally on the OAK device and not on a cloud server, it’s still edge computing.

Cloud computing only applies if the actual computation (inference) is happening on a remote cloud server.

Thanks,
Jaka

michaelsalerno

@jakaskerl

Thank you for all your help!

One last question,

What NN is it using? Luxonis published information on latency that I would like to use, and I am curious to know which one it is so I can accurately talk about latency.