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

Hi Tinbum, thanks for this I can confirm this version is working and I did not detect any malicious activity. (For others doesn't mean this is safe)
As it's not the version referenced in the Official Manual or Doco, do you have any background on where this version came from?

I do note a bug, where Mode = Low performs better than omitting the mode or putting Mode = High, other than that the Windows GPU performance is good.
I am using an old GTX 745 with no display connected (although this overclocks very high, just running stock) ~66ms to process 180x720 resolution images, compared to 300ms with an i7-6700.

Using on-board graphics for the display, to avoid conflicting with BI QuickSync.
 
a bug, where Mode = Low performs better than omitting the mode or putting Mode = High

by perform better, do you mean it has a faster response time?

omitting the mode likely defaults to the medium accuracy setting. I think 'low' (accuracy) should be fastest, followed by default, and then high (accuracy).
 
by perform better, do you mean it has a faster response time?

omitting the mode likely defaults to the medium accuracy setting. I think 'low' (accuracy) should be fastest, followed by default, and then high (accuracy).
Yes, by better I meant faster.
Thankyou for clarifying what mode meant, I incorrectly believed it was performance not accuracy.

GPU bursts to 2% usage, so will try your multi-deepstack instance suggestion.
 
  • Like
Reactions: cjowers
As it's not the version referenced in the Official Manual or Doco, do you have any background on where this version came from?

I do note a bug, where Mode = Low

It's official Deepstack, that's not a bug it will be faster in low.
 
Last edited:
Still having major issues on the person identification side. Constantly missing things as obvious as this.
1610372029275.png
1610372048800.png1610372085636.png



Cool down period was not an issue (it picked up the person walking up on my driveway camera, but clicked only 2 pics in 4 minutes for this food delivery guy, and both times picked up the potted plants only.

Also the driveway cam while it caught the person walking from his car towards my door, didn't capture him when he left the door and went back to his car a few minutes later, I assume that's more of a BI settings issue, though not sure why it captures the person on the way in but not the way out?
 
When using the latest VorlonCD / bi-aidetection

In the log sometimes I see:

"Could not trigger URL xx.xx.xx.xx"

It looks like this only happens when it is already triggered. Am I correct in assuming it will throw this error while it is already triggered or should I look further into what is happening in my system?
 
Anyone that is running the 2020.12 beta of deepstack, any clue what this is causing the below error? this is what is coming of in the stderr.txt file ( I am running the CPU version)

Here is the error:
Process Process-1:
Traceback (most recent call last):
File "D:\obj\windows-release\37amd64_Release\msi_python\zip_amd64\process.py", line 297, in _bootstrap
File "D:\obj\windows-release\37amd64_Release\msi_python\zip_amd64\process.py", line 99, in run
File "C:\DeepStack\intelligencelayer\shared\detection.py", line 138, in objectdetection
os.remove(img_path)
FileNotFoundError: [WinError 2] The system cannot find the file specified: 'C:\\Users\\baluc\\AppData\\Local\\Temp\\DeepStack\\f28c9002-abb6-4e2e-a531-e7aa45b48e81'
 
FileNotFoundError: [WinError 2] The system cannot find the file specified: 'C:\\Users\\baluc\\AppData\\Local\\Temp\\DeepStack\\f28c9002-abb6-4e2e-a531-e7aa45b48e81'
that looks like the deepstack server api key file.
edit- on 2nd thought, maybe that's just a standard windows temp filename format... i got no idea

is the file there? permission / access issue maybe? are you running everything as admin?
does it run at all?
 
Last edited:
Hi, I just installed Blue Iris version 5.3.7.12 (demo mode) and I am trying to configure this. AI Tool won't recognize DeepStack. I have tried to manually run it in powershell and start it from the "DeepStack" tab in AI Tool. DeepStack version: DeepStack-Installer-CPU.2021.01.beta.exe.

OnGuard works, it recognizes DeepStack started from powershell and the DeepStack web page is also running. I have tried restarting the computer. Server.exe, python.exe or redis-server.exe are not running when I start DeepStack. I can't install it in docker because I need to enable Hyper-V in bios and I can't access the computer right now.

1610454283588.png

We don't need the DeepStack signup (api key) anymore and the beta version is their main download link in docs Using DeepStack with Windows 10 (CPU and GPU) | DeepStack v1.1.2 documentation
 
Last edited:
Hi, I just installed Blue Iris version 5.3.7.12 (demo mode) and I am trying to configure this. AI Tool won't recognize DeepStack. I have tried to manually run it in powershell and start it from the "DeepStack" tab in AI Tool. DeepStack version: DeepStack-Installer-CPU.2021.01.beta.exe.

OnGuard works, it recognizes DeepStack started from powershell and the DeepStack web page is also running. I have tried restarting the computer. Server.exe, python.exe or redis-server.exe are not running when I start DeepStack. I can't install it in docker because I need to enable Hyper-V in bios and I can't access the computer right now.

View attachment 79473

We don't need the DeepStack signup (api key) anymore and the beta version is their main download link in docs Using DeepStack with Windows 10 (CPU and GPU) | DeepStack v1.1.2 documentation

AI Tools is looking for C:\DeepStack\server\server.exe The latest version of DeepStack changed this executable to deepstack.exe. Just rename it back to server.exe. That's it!
 
  • Like
Reactions: Quru
that looks like the deepstack server api key file.
edit- on 2nd thought, maybe that's just a standard windows temp filename format... i got no idea

is the file there? permission / access issue maybe? are you running everything as admin?
does it run at all?
It's running but deepstack errors out after a short while. Found the newest beta of deepstack last evening and installed it along with the newest build of AI Tool it has been running overnight, will check it later and see if that fixes the issue.
 
AI Tools is looking for C:\DeepStack\server\server.exe The latest version of DeepStack changed this executable to deepstack.exe. Just rename it back to server.exe. That's it!

Actually this didn't fix the problem, still errors in AI Tools. I managed to enable Hyper-V without going to BIOS, I love Lenovo's tools. DeepStack Docker version now works with AI Tools, but I have performance issues with three 2560x1920 H265 cameras. I have I5-6400 with 8GB ram, I think I need more ram or it is because of the Demo version. I bought license, but still haven't got it :nervous:. It only sometimes captures the video when triggering, I have no idea whats going on. I configured settings using first post recommended settings, I only changed the trigger url to this.
 
It's running but deepstack errors out after a short while. Found the newest beta of deepstack last evening and installed it along with the newest build of AI Tool it has been running overnight, will check it later and see if that fixes the issue.
Replaced both deepstack and AI-TOOl with the latest versions and that took care of it. As I was perusing GitHub I see my problem was a known issue.
 
Any idea why items would get added to queue and never get out of the queue unless I close and re-open AI Tool? I have tried older and new version of VorlonCD with the same results. I checked through the issues on Git and didn't find anything similar.

Deepstack is hosted on docker and continues running without issues.

[13:55:05.931]: WaitForNextURL> ---- #1: Re-added known URL: [13:55:05.932]: WaitForNextURL> ...Found 1 AI URL's in settings.
[13:55:06.136]: OnCreatedAsync>
[13:55:06.137]: OnCreatedAsync> ====================== Adding new image to queue (Count=1): G:\aiinput\driveway.20210112_135506134.jpg
[13:55:08.137]: OnCreatedAsync>
[13:55:08.138]: OnCreatedAsync> ====================== Adding new image to queue (Count=2): G:\aiinput\driveway.20210112_135508134.jpg
..
[14:02:16.051]: OnCreatedAsync> ====================== Adding new image to queue (Count=45): G:\aiinput\driveway.20210112_140216048.jpg
[14:02:18.115]: OnCreatedAsync>
[14:02:18.117]: OnCreatedAsync> ====================== Adding new image to queue (Count=46): G:\aiinput\driveway.20210112_140218113.jpg
 
For those running the Windows GPU I noticed Deepstack only uses about 1% of the GPU processing power so this could be of interest if you run the VorlonCD version of AITool. Note AITool will only start one instance at the moment.

Can confirm this is working well with the version you supplied with 6 instances. Going beyond 6 works, but find I hit the Deepstack idle issue, ie: Deepstack seems to pause for 200-300ms, if it goes quiet for a while.

Had to launch from powershell, launching multiple instances from the CMD seemed to conflict with each other (ofcourse different ports were used)
 
Is my driveway camera placement wrong? Seems like DS/AITOOL is not picking up persons while near or in the same frame as my car.

1st shot picked me up just fine
Screen Shot 2021-01-12 at 9.37.24 PM.png
Then I walked to my passenger door and it didn't pick me up
Screen Shot 2021-01-12 at 9.37.33 PM.png
Then it picked me up again as I was walking away
Screen Shot 2021-01-12 at 9.38.16 PM.png

Should I move the camera to the center of the garage?
Thanks.
 

Attachments

  • Screen Shot 2021-01-12 at 9.38.07 PM.png
    Screen Shot 2021-01-12 at 9.38.07 PM.png
    731.1 KB · Views: 35
  • Like
Reactions: spammenotinoz
Is my driveway camera placement wrong? Seems like DS/AITOOL is not picking up persons while near or in the same frame as my car.

1st shot picked me up just fine
View attachment 79529
Then I walked to my passenger door and it didn't pick me up
View attachment 79528
Then it picked me up again as I was walking away
View attachment 79526

Should I move the camera to the center of the garage?
Thanks.
Camera placement is not bad, but this is a Good reference case, on why an areas should ideally be covered by 2 cameras with different views. Specifically a second cam on the left pointing to the right will get faces of people approaching from the right and detect a person.
Recognizing a person 3/4 hidden behind a car and with such low contrast will be a tough ask.
See if you can improve the contrast, but think you will struggle unless you can get the cam to see more torso.
Also highlights the importance of constantly recording, that way you just need to detect the person approaching or leaving and then review footage in between.
 
Camera placement is not bad, but this is a Good reference case, on why an areas should ideally be covered by 2 cameras with different views. Specifically a second cam on the left pointing to the right will get faces of people approaching from the right and detect a person.
Recognizing a person 3/4 hidden behind a car and with such low contrast will be a tough ask.
See if you can improve the contrast, but think you will struggle unless you can get the cam to see more torso.
Also highlights the importance of constantly recording, that way you just need to detect the person approaching or leaving and then review footage in between.
Yeah, I have 24x7 recording of the substream. Might have to shell out some cash and buy another camera though, just have to convince the wifey.