Shinobi Video DeepStacks plugins

Hi all,

Created 2 new plugins (unofficial yet) for Shinobi Video that uses DeepStack Face Recognition and Object Detection.
Available at:

Worked up until 2 months ago with BI,
Once added DeepStack integration it stopped working as it worked before, switched to Shinobi Video.

Plugins and now it covers all my camera (16), works with much less GPU consumption than other plugins for object detection (for face didn’t work for me), response of less than 50 msec.

In addition, added HA integration for Shinobi Video that leverges the 2 plugins - available at HACS - look for Shinobi Video NVR

1 Like This is great work. Thanks for sharing this. We will feature this on the DeepStack Dev Center

In case you are not aware, support for DeepStack is now officially added to Blue Iris latest version and you can learn more about setting this up in this video.

Also, are you going to provide visual images of the plugins at work in the repositories or publish a tutorial/guide on the integration? It will be good to have any of both. :grinning:

hi, I switched to Shinobi Video since BI became less stable for me, after playing with it a bit, I found that it works for me better,
Instead of 2 servers (with GPUs) of Windows 10 and Ubuntu Server, I have just 1 (Ubuntu) now, which is easier to maintain (all the 26 containers of home automation in a single docker compose).

With Shinobi Video I’m able to use face-reco (DeepStack) for all camera which work more efficiant (less network traffic) than BI when I used 2 camera for face-reco.

As for the documentation, I extended it earlier today to be in the same format as other plugins of Shinobi Video,
Also asked in Shinobi Video discord’s dev channel to assist me to make it an official plugin,
It seems that Moe Alam (founder of it) really interested in and installed it, after several hours reused the plugin for License Plate model (the custom one from dev portal of DeepStack), hope it will become official soon.

I will work on a full guide later this week and post it here, thanks again for the amazing product!


1 Like

Awesome. Well done and thanks for sharing this publication . We will share widely in our publications as well.