DrewBarfield

I think installing depthai_experiments causes something similar too. (Or maybe the requirements for a single experiment.)

Here's the output trying to run depthai_hand_tracker after installing depthai_experiments:

XXXXX@XXXXX:/workdrive/Repositories/oak-d-lite/depthai_hand_tracker$ python3 demo.py -e -g --lm_model lite
Palm detection blob     : /workdrive/Repositories/oak-d-lite/depthai_hand_tracker/models/palm_detection_sh4.blob
Landmark blob           : /workdrive/Repositories/oak-d-lite/depthai_hand_tracker/models/hand_landmark_lite_sh4.blob
PD post processing blob : /workdrive/Repositories/oak-d-lite/depthai_hand_tracker/custom_models/PDPostProcessing_top2_sh1.blob
Sensor resolution: (1920, 1080)
Internal camera FPS set to: 36
Internal camera image size: 1152 x 648 - pad_h: 252
Creating pipeline...
Creating Color Camera...
Creating Palm Detection pre processing image manip...
Creating Palm Detection Neural Network...
Creating Palm Detection post processing Neural Network...
Creating Hand Landmark pre processing image manip...
Creating Hand Landmark Neural Network (2 threads)...
Pipeline created.
[184430101116701200] [7.841] [NeuralNetwork(4)] [warning] Network compiled for 4 shaves, maximum available 13, compiling for 6 shaves likely will yield in better performance
Pipeline started - USB speed: HIGH
[184430101116701200] [8.073] [NeuralNetwork(8)] [warning] Network compiled for 4 shaves, maximum available 13, compiling for 6 shaves likely will yield in better performance
[184430101116701200] [8.085] [NeuralNetwork(4)] [warning] The issued warnings are orientative, based on optimal settings for a single network, if multiple networks are running in parallel the optimal settings may vary
[184430101116701200] [8.085] [system] [error] Attempted to start Color camera - NOT detected!
[184430101116701200] [8.085] [NeuralNetwork(8)] [warning] The issued warnings are orientative, based on optimal settings for a single network, if multiple networks are running in parallel the optimal settings may vary
^CTraceback (most recent call last):
  File "demo.py", line 85, in <module>
    frame, hands, bag = tracker.next_frame()
  File "/workdrive/Repositories/oak-d-lite/depthai_hand_tracker/HandTrackerEdge.py", line 462, in next_frame
    in_video = self.q_video.get()
KeyboardInterrupt

...and running the depthai_demo.py:

XXXXX@XXXXX:/workdrive/Repositories/oak-d-lite/depthai$ python3 depthai_demo.py 
Using depthai module from:  /home/drew/.local/lib/python3.8/site-packages/depthai.cpython-38-x86_64-linux-gnu.so
Depthai version installed:  2.11.1.0
Version mismatch between installed depthai lib and the required one by the script. 
                Required:  2.13.3.0
                Installed: 2.11.1.0
                Run: python3 install_requirements.py 

Disconnecting for 30 seconds and reinstalling the requirements for the depthai repo resolved my problem:
python3 -m pip install -r requirements.txt

I've heard of Project North Star. I didn't realize that it was still on-going. I'll definitely check them out again.

Thanks @erik !

What if the "pole" was horizontal as in the case of a balcony handrail?

The VESA angle is interesting. There are a lot of options. My only concern would be outdoor/weatherproof options.

Oh, yeah. I've been meaning to ask: Do acrylic domes have any negative effect on these cameras? Any chance they could increase false positives, reflections, etc.

Amazon link

@Brandon / @erik,

I hope this makes some sense. The image did not come out exactly as planned. However, it should convey the basic idea. The back plate screws to a wall stud. A hole must be drilled into the wall for the female Ethernet connector and cable. The camera sits on the male connector and a screw can be threaded into the ISO 1222:2010 camera threaded bore.

How can I display the output (a preview or video) as an inline video in a Jupyter Notebook? The 4K RGB MobileNetSSD sample works just fine with the external windows, but I want to access JupyterLab across the network. In order for this to work properly, I'll need the video to display in the notebook. Otherwise the notebook server will be showing a window I can't see or close.

I'm also wondering if there are other best practices for DepthAI in JupyterLab. For example, you should always explicitly clean up your code, or your script will hang and crash the kernel. See the last 3 lines:

import depthai as dai
pipeline = dai.Pipeline()

camRgb = pipeline.createColorCamera()
camRgb.setPreviewSize(300, 300)
camRgb.setBoardSocket(dai.CameraBoardSocket.RGB)
camRgb.setResolution(dai.ColorCameraProperties.SensorResolution.THE_1080_P)
camRgb.setInterleaved(False)
camRgb.setColorOrder(dai.ColorCameraProperties.ColorOrder.RGB)

xoutRgb = pipeline.createXLinkOut()
xoutRgb.setStreamName("rgb")
camRgb.preview.link(xoutRgb.input)

with dai.Device(pipeline) as device:
    device.startPipeline()
    qRgb = device.getOutputQueue(name="rgb", maxSize=4, blocking=False)
    print('Press \'Q\' to Quit')
    while True:
        inRgb = qRgb.get()
        cv2.imshow("bgr", inRgb.getCvFrame())
        if cv2.waitKey(1) == ord('q'):
            break
    device.close()
    
cv2.destroyAllWindows()
del pipeline

Any tips are greatly appreciated!
-Drew