DeepStack on Rasperry PI makes it easier to develop and deploy embedded smart applications. This is a lighter version but with full support for all of deepstack's features including Object Detection, Face APIs, Scene Recognition.
Custom ONNX, Tensorflow, Keras and Openvino models support will be included in the BETA version.
Note that due to the limited compute capacity of the Raspberry PI, DeepStack requires an Intel Neural Compute Stick (NCS)
The PI Version is in alpha and would be regularly improved for optimal performance.
Minimum System Requirements
Raspberry PI 3B+
Intel Movidius Neural Compute Stick 2
wget https://deepstackfiles.s3.us-east-2.amazonaws.com/pi/install-deepstack.shsh install-deepstack.sh
DeepStack on the raspberry pi depends on python3.5 or Python3.7 installed.
Due to a number of factors, running DeepStack on the PI is slightly different.
Below are the various ways to run deepstack. Ensure the NCS is plugged in before running deepstack
sudo deepstack start "VISION-SCENE=True"
sudo deepstack start "VISION-DETECTION=True"
sudo deepstack start "VISION-FACE=True"
The first time you run DeepStack, you need to activate it following the process below.
Once you initiate the run command above, visit localhost:80 in your browser. The interface below will appear.
Get an activation key from DeepStack.cc , paste it in the box and click Activate Now. Once activated, the interface below will appear.
Once installed, run the example scene recognition code to verify your installation is working
sudo deepstack stop
DeepStack by default runs on port
80 you can run it on your choice port using the syntax below.
sudo deepstack start "VISION-SCENE=True PORT=8080"
Only one api type can be run at once, for example, you would encounter errors if you attempt to run both
Object Detection and
Face at the same time.
Custom Models are not supported at present, support for this would be added in the BETA release.
Only One instance of DeepStack can be run at present, running a new instance would close the other , this is enforced due to limitations of the Neural Compute Stick.
Use of Multiple NCSs is not yet supported. Only one can be used in this release.
DeepStack Not Responding : This might occur if you start DeepStack without the NCS plugged in or you sometimes when you stop and rerun DeepStack. In any case, simply restart your Raspberry PI, plug in the NCS and start deepstack.
Invalid Image : This can occur if you send a very high resolution image, about 2000px * 2000px, to the
object detection or