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

[aesterling]

can you provide more detail on the "copy/save settings" part? I added the &jpg=[imagepath] and checked merge annotations into images, but not sure about the copy. I see copy alert images to folder... is that the feature ? do i have to copy the image to a new image? if so, do hard code or can i use [imagepath] variable? thanks

If you want BI to use the image that was analyzed by AI Tool/Deepstack as the alert image, then I believe you can just use the [ImagePath] variable. If you want BI to use the annotated image that includes the box around detected objects and text label, then enabled AI Tool's "Merge Annotations into images" and "Copy Alert images to folder" options and point the URL trigger "&jpeg=path" to that path instead. You can click the green text at the top of the Actions > Settings window to show you what the variable will output. If I got any of that wrong, I'm sure @Chris Dodge could correct me.

 

[Tinbum]

For the odd display issues, make a shortcut to AITOOLS somewhere, right-click > properties > compatibility tab > change high dpi settings, and see if they help for the next time you start it. Or maybe edit aitools.exe.config and set the value for DpiAwareness to "false" rather than "PerMonitorV2".

That worked for me on a laptop.

 

[charredchar]

Mine has 4 GB of RAM. Not sure how it would perform with only 1 GB. I can check my memory usage later today and let you know.


Sent from my iPhone using Tapatalk

Thank you! Is there any possible way you could limit your DeepStack install to under 1GB to simulate what it would be like so I could get an idea of what the response time would be? If it is too much of a hassle, don't bother. I am just worried that I go purchase a NCS2 and find out it ends up worse on my Pi4 than the VM.

 

[Tinbum]

Thank you! Is there any possible way you could limit your DeepStack install to under 1GB to simulate what it would be like so I could get an idea of what the response time would be? If it is too much of a hassle, don't bother. I am just worried that I go purchase a NCS2 and find out it ends up worse on my Pi4 than the VM.

Isn't the PI4 2Gb minimum. My PI 4 is faster than my PI3b+ by about 1/3 at least

 

[charredchar]

Isn't the PI4 2Gb minimum. My PI 4 is faster than my PI3b+ by about 1/3 at least

No, 1GB models exist, they are basically the lowest model though. Considering I was previously using a Pi with 512MB of RAM I figured the 1GB would be more than enough for most projects I'd use it. DeepStack really requires quite a lot and not really something I'd ever consider for a Pi if my VM wasn't having so much trouble and/or an NCS2 would work on PC.

 

[WildDoktor]

I got the 1.77 1.81 version of aitool-vorloncd setup and running, but I'm seeing something that seems wierd to me: "very high" power usage, and it's using 46.7% of my CPU. Is this right? This is happening when even when BI cpu usage is 0%.

 

[Tinbum]

No there is something wrong there. Mine uses up to about 3% ish processing 60+ images a minute.
Whats the (9) you haven't got 9 instances running have you?

 

[hazzard8]

I'm running DeepquestAI (via Docker on Windows) on an i3-7100U with 8gb of RAM. I can see that most of you have a detection time well below 1 second. Mine is around 2 seconds:



I've tried to run DeepquestAI on my i7-4790 with 16gb RAM, but still got around 2 seconds. Is there anything I can do to try to reduce detection time? I've got an Nvidia RTX2060, but running the nvidia-option did not help.

 

[balucanb]

Is anyone else having these issues? Am I missing something? I was running the latest version of AI-Tool from gentlepumpkin with no major issues, deleted all the files and downloaded AITOOL-VORLONCD V1.81 and am now running this. I have 6 cameras and am running everything on a Win10 pro machine with a i7-4790 @3.6 and 32GB of ram, 64 bit OS. When I review the footage in BI I see events that should be causing alerts, i.e. cars driving down the road in front of the house. I am not using any masks, limits are set to 0/100% pre triggers are either 5 or 10 secs. and BI is supposed to be taking stills every 5 secs. All the stills go into 1 folder and I am not monitoring any others, When I check the folder I am only seeing pics from a few of the cameras, when I check the log in AI-Tool I only see pics from a single camera? Also have alert pics being sent to Telegram. With the prior set up I was getting almost to many alerts (busy street) now as I have stated I am getting the opposite. If I walk around and stand in front of each camera it will alert however? Any clue what is going on?

 

[Tinbum]

Is anyone else having these issues? Am I missing something? I was running the latest version of AI-Tool from gentlepumpkin with no major issues, deleted all the files and downloaded AITOOL-VORLONCD V1.81 and am now running this. I have 6 cameras and am running everything on a Win10 pro machine with a i7-4790 @3.6 and 32GB of ram, 64 bit OS. When I review the footage in BI I see events that should be causing alerts, i.e. cars driving down the road in front of the house. I am not using any masks, limits are set to 0/100% pre triggers are either 5 or 10 secs. and BI is supposed to be taking stills every 5 secs. All the stills go into 1 folder and I am not monitoring any others, When I check the folder I am only seeing pics from a few of the cameras, when I check the log in AI-Tool I only see pics from a single camera? Also have alert pics being sent to Telegram. With the prior set up I was getting almost to many alerts (busy street) now as I have stated I am getting the opposite. If I walk around and stand in front of each camera it will alert however? Any clue what is going on?

If the images are not in the folder that AITool is monitoring then it looks like the set up in BI is wrong.

 

[charredchar]

Is anyone else having these issues? Am I missing something? I was running the latest version of AI-Tool from gentlepumpkin with no major issues, deleted all the files and downloaded AITOOL-VORLONCD V1.81 and am now running this. I have 6 cameras and am running everything on a Win10 pro machine with a i7-4790 @3.6 and 32GB of ram, 64 bit OS. When I review the footage in BI I see events that should be causing alerts, i.e. cars driving down the road in front of the house. I am not using any masks, limits are set to 0/100% pre triggers are either 5 or 10 secs. and BI is supposed to be taking stills every 5 secs. All the stills go into 1 folder and I am not monitoring any others, When I check the folder I am only seeing pics from a few of the cameras, when I check the log in AI-Tool I only see pics from a single camera? Also have alert pics being sent to Telegram. With the prior set up I was getting almost to many alerts (busy street) now as I have stated I am getting the opposite. If I walk around and stand in front of each camera it will alert however? Any clue what is going on?

As Tinbum said, missing images in the folder is an issue with Blue Iris, double check your settings (Record, Trigger, etc) then manual trigger to ensure they end up there.
As for AITool, if images aren't showing up in the history tab then double check the AITool settings to ensure you have all cameras added properly. If the name of a camera is wrong, for instance, it won't snatch those images from that folder, it'll ignore them.

 

[balucanb]

Sorry was not clear enough, they are in a monitored in a folder and it is the one AI-Tool is looking at. I am not monitoring any other folders than this one - no entry in the input folder block and monitor subfolders is not checked.

 

[balucanb]

As Tinbum said, missing images in the folder is an issue with Blue Iris, double check your settings (Record, Trigger, etc) then manual trigger to ensure they end up there.
As for AITool, if images aren't showing up in the history tab then double check the AITool settings to ensure you have all cameras added properly. If the name of a camera is wrong, for instance, it won't snatch those images from that folder, it'll ignore them.

Sorry was not clear enough, they are in a monitored in a folder and it is the one AI-Tool is looking at. I am not monitoring any other folders than this one - no entry in the input folder block and monitor subfolders is not checked.

 

[Tinbum]

You say you deleted AITool then downloaded the new version. Did you keep your camera folder and settings. Have you set up new cameras or are they imported from your old AITool?

 

[balucanb]

You say you deleted AITool then downloaded the new version. Did you keep your camera folder and settings. Have you set up new cameras or are they imported from your old AITool?

No didn't pay attention to what was probably the best way to upgrade, I deleted all of it and the new version pulled in all my cameras from BI

 

[Tinbum]

I'm not sure how that way works so can't really help.

 

[charredchar]

Sorry was not clear enough, they are in a monitored in a folder and it is the one AI-Tool is looking at. I am not monitoring any other folders than this one - no entry in the input folder block and monitor subfolders is not checked.

Not to sound rude but I told you where to look to fix your issues. Check your settings.

 

[balucanb]

Not to sound rude but I told you where to look to fix your issues. Check your settings.

LOL- No worries here! I'm the king of rude and I try. LOL. I did not think you were rude BTW. I didn't even realize I was replying to a different person until you re-posted. All your suggestions are valid but, all you mentioned are fine. HOWEVER!!.... As is the case 99% of the time (with me) the problem was the 12 inches of space between my forehead and keyboard Your suspicions were valid, I rechecked some of the "detection zone" sliders and of course they were way out of whack, adjusted them and it seems to be doing better now!

 

[balucanb]

Our new mod of AITOOL lets you specify as many DeepStack URL's as you like in settings. If you have more than one image in the queue, they will be processed in parallel in each deepstack server.
The log does get a little more complicated to view since results from each URL are mixed together, but it seems to work well from my initial testing. (@GentlePumpkin )

Releases · VorlonCD/bi-aidetection

Alarm system for Blue Iris based on Artificial Intelligence. - VorlonCD/bi-aidetection

github.com

Sorry for what I am sure is a simple answer but this is all new to me. If your set up is not top notch (read- my system is slow) would this help speed things up? also in order to have more than 1 DS URL do you have to use one of the paid plans? Again sorry all new to me, no clue how I would get multiple URL's. TIA.

 

[Tinbum]

Sorry for what I am sure is a simple answer but this is all new to me. If your set up is not top notch (read- my system is slow) would this help speed things up? also in order to have more than 1 DS URL do you have to use one of the paid plans? Again sorry all new to me, no clue how I would get multiple URL's. TIA.

You dont have to pay- just set another instance of deepstack up on a Pi or another computer.