Hello, I was following this [tutorial](https://docs.luxonis.com/en/latest/pages/tutorials/deploying-custom-model/ ) to deploy a customed model and everything seem to work fine, but at the end, the camera doesn't show up on screen and the code is stuck. It looks like it doesn't receive msgs, so never enter the if condition: [URL=https://i.imgur.com/DrtasrD.png][IMG]https://i.imgur.com/DrtasrDh.png[/IMG][/URL] [URL=https://i.imgur.com/YsSYcNT.png][IMG]https://i.imgur.com/YsSYcNT.png[/IMG][/URL] But the camera is working fine with the yunet code from[ here](https://github.com/luxonis/depthai-experiments/tree/master/gen2-face-detection), I just can't figure out what's wrong in the main of the one from deploying custom model. Thank you

Hi @"Touk"#2302 Looks like the syncing isn't working properly, so no new (synced) messages are shown. Try removing the syncing part to see if the stream works. Alternatively you can try with the [Sync](https://docs.luxonis.com/projects/api/en/latest/components/nodes/sync_node/) node, which abstracts the syncing part and creates less room for error. Thanks, Jaka

Hi @"Touk"#2302 Could be a link missing perhaps? Try visualizing the pipeline with https://github.com/luxonis/depthai_pipeline_graph. Thanks, Jaka

Depthai main code not working

Touk

Hello,

I was following this tutorial to deploy a customed model and everything seem to work fine, but at the end, the camera doesn't show up on screen and the code is stuck. It looks like it doesn't receive msgs, so never enter the if condition:

But the camera is working fine with the yunet code from here, I just can't figure out what's wrong in the main of the one from deploying custom model.

Thank you

jakaskerl

Hi @Touk
Looks like the syncing isn't working properly, so no new (synced) messages are shown. Try removing the syncing part to see if the stream works. Alternatively you can try with the Sync node, which abstracts the syncing part and creates less room for error.

Thanks,
Jaka

Touk

Hello,
Thank you for your help, but so far, it is still not working.
Here below is how I removed the sync part (maybe it's wrong), but I don't have error, just nothing from the camera. Can it be because I am using a USB2 (USB3 for camera to USB2 for my computer) ?
with dai.Device() as device:

stereo = 1 < len(device.getConnectedCameras())

device.startPipeline(create_pipeline(stereo))

queues = {}

# Create output queues

for name in ["color", "detection", "recognition"]:

    queues[name] = device.getOutputQueue(name)

while True:

    for name, q in queues.items():

        if q.has():

            frame = None

            detections = None

            recognitions = None

            if name == "color":

                frame = q.get().getCvFrame()

            elif name == "detection":

                detections = q.get().detections

            elif name == "recognition":

                recognitions = q.get()

            

            if detections is not None:

                for detection in detections:

                    bbox = frame_norm(frame, (detection.xmin, detection.ymin, detection.xmax, detection.ymax))

                    cv2.rectangle(frame, (bbox[0], bbox[1]), (bbox[2], bbox[3]), (10, 245, 10), 2)

                    cv2.putText(frame, "Age: {}, Gender: {}".format(age, gender_str), (bbox[0], bbox[1] + 30), cv2.FONT_HERSHEY_TRIPLEX, 1.5, (0, 0, 0), 8)

                    cv2.putText(frame, "Age: {}, Gender: {}".format(age, gender_str), (bbox[0], bbox[1] + 30), cv2.FONT_HERSHEY_TRIPLEX, 1.5, (255, 255, 255), 2)

                cv2.imshow("Camera", frame)

            elif recognitions is not None:

                rec = recognitions.getFirstLayerFp16()

                index = np.argmax(rec, axis=1)

                text = "No Detection"

                color = (0,0,255) # Red

                if index is not None:

                    text = index

                    color = (0,255,0)

                cv2.putText(frame, text, (10, 50), cv2.FONT_HERSHEY_TRIPLEX, 1.5, color, 2)

jakaskerl

Hi @Touk

Touk Can it be because I am using a USB2 (USB3 for camera to USB2 for my computer) ?

Most likely not.
Which part blocks? Maybe just try with:

colorQueue.get()
print("got color")
detection.get()
print("got detection")
.
.
.

Thanks,
Jaka

Touk

You're right, it actually gets stuck when it tries to get the detection. I don't know why though. I changed nothing on the detection part, it is exactly what I got from the tutorial.
Recognition get stuck as well.

jakaskerl

Hi @Touk
Could be a link missing perhaps? Try visualizing the pipeline with luxonis/depthai_pipeline_graph.

Thanks,
Jaka