[tool] [tutorial] Free AI Person Detection for Blue Iris

1903? And DeepQuestAI runs fine?
Can you plz do a right click on start then run then type in winver and let me know what version it shows?
Thanks
 
Hmm strange when I install DeepQuestAi it always gives me these errors I mentioned before ... thought it was DeepQuestAI ..but now it runs fine for you...might start a new virtual server and try again
 
That I haven't figured it out yet (start deepstack as a service or something). Still running my Docker - upgrades automatically (via watchtower, auto start etc).
Maybe docker on separate server is better but I just want to see the speeds I can get if it was local.
Yes all ports etc are OK - since it works fine with my docker deepstack with same port I am using on Windows I am 100% is not an AI tool or deepstack startup config issue
 
So far so good... Getting to grips with it and testing it

iEBs2Ts.png


Do we have to have the 1 minute cool downs? Doesn't seem to work for me without a minute in there.

SXwQYk6.png
 
Tinbum said:
Works ok for me with 0 mins.
Click to expand...

Hmm need to test more, some times there's false negatives but it could be due to my camera placement. I'm waiting for my new 4K camera which I'll mount much lower down which should help with that.
 
I'm getting false positives but I'm getting there. Just like to sort how it will run automatically on computer startup. I also want to be able to detect cars when there are cars parked in the field of view. Would be nice to be able to minimize the server screen. Still learning.
 
Tinbum said:
I'm getting false positives but I'm getting there. Just like to sort how it will run automatically on computer startup. I also want to be able to detect cars when there are cars parked in the field of view. Would be nice to be able to minimize the server screen. Still learning.
Click to expand...

You're getting false positives from the AI? In terms of making it auto run, should be able to with a script. But might be nice to have it all integrated into AI tool.
 
Yes had some saying people in them but I can't see them. !! Never written a script. I do a little bit of Arduino programming but that's it.
 
Have many cameras do you guys have? I noticed in BI that anything below 6 cameras works really well when these cameras are cloned. I have 8 cameras. I as soon as I clone all of them weird things start to happen.

First of all the * next tot the camera name (my understanding about the * is that it shows when a camera has a clone) moves around a bit after each reboot. At least on 2 cloned camera the * is shows and the rest of the 6 cameras shows the *. And is random - different 2 cloned will have the * after each reboot.

Second - motion detection on cloned cameras will work well EXCEPT for the 2 cloned cameras showing the * If I reboot then the 2 will work well but the next 2 with * will not. I can stand in front of the cloned cameras showing the * for a number of minutes before I get an alert. Looking at the DeppStackAI docker logs shows that the picture is processed under 1s (normally around 800ms) and the alert come to me pretty much instantly AFTER the cloned camera with the * decides to alert.

Do you guys have more than 6 cameras and experience anything like that? Maybe your false positives are about BI and cloned cameras and not about the AI Tool?
 
Last edited:
MnM said:
Have many cameras do you guys have? I noticed in BI that anything below 6 cameras works really well when these cameras are cloned. I have 8 cameras. I as soon as I clone all of them weird things start to happen.

First of all the * next tot the camera name (my understanding about the * is that it shows when a camera has a clone) moves around a bit after each reboot. At least on 2 cloned camera the * is shows and the rest of the 6 cameras shows the *. And is random - different 2 cloned will have the * after each reboot.

Second - motion detection on cloned cameras will work well EXCEPT for the 2 cloned cameras showing the * If I reboot then the 2 will work well but the next 2 with * will not. I can stand in front of the cloned cameras showing the * for a number of minutes before I get an alert. Looking at the DeppStackAI docker logs shows that the picture is processed under 1s (normally around 800ms) and the alert come to me pretty much instantly AFTER the cloned camera with the * decides to alert.

Do you guys have more than 6 cameras and experience anything like that? Maybe your false positives are about BI and cloned cameras and not about the AI Tool?
Click to expand...

I've only got two cameras at the moment, one of them cloned. Got some 4K turrets on the way.

If this is a genuine issue, report it to to BI Support. And s what they say. I've been rather disappointed with BI5 performance, it was meant to run much better. But it runs the same as V4 for me.
 
I've 40 odd cameras on mine but I haven't cloned any of them. At the moment I'm just using this to send me notifications.
 
No not business just live on an old farm. Cameras are mixture. A few Panasonic and a lot of cheap Chinese ones. Computer is Intel I7 8700K with about 5 ssd hard drives All works really well. It's set pretty sensitively.

Stats at the moment are 155 alerts, 1753 irrelevant and 947 false. It keeps thinking the horse is a person though!!
 
Tinbum said:
No not business just live on an old farm. Cameras are mixture. A few Panasonic and a lot of cheap Chinese ones. Computer is Intel I7 8700K with about 5 ssd hard drives All works really well. It's set pretty sensitively.

Stats at the moment are 155 alerts, 1753 irrelevant and 947 false. It keeps thinking the horse is a person though!!
Click to expand...

Unfortunately I can't claim to have any horses to test, would love to one day when I move out London. XD

The AI definitely needs work some improvement but it's better than getting so many false positives IMHO. There's a few things that it's missed that I'm trying to get to the bottom of.
 
You must log in or register to reply here.
Top Bottom