@"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?

@"filipjeretina"#2315 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.

@"michaelsalerno"#p26131 > @"michaelsalerno"#p26131 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

Dear @"jakaskerl"#1112 , I have tried it on my end, and i get an error stating that it can't connect to the API. From Michael

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

@"jakaskerl"#1112 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"#p26206 Can you paste this error you are getting? > @"michaelsalerno"#p26207 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

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( File "C:\\Users\\sale4088\\AppData\\Local\\Packages\\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\\LocalCache\\local-packages\\Python310\\site-packages\\urllib3\\connectionpool.py", line 404, in _make_request self._validate_conn(conn) File "C:\\Users\\sale4088\\AppData\\Local\\Packages\\PythonSoftwareFoundation.Python.3.10_qbz5n2kfra8p0\\LocalCache\\local-packages\\Python310\\site-packages\\urllib3\\connectionpool.py", line 1060, in _validate_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: : 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(': 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 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) 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 231, in _parse_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(': Failed to establish a new connection: [Errno 11001] getaddrinfo failed'))

No service use

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.