CodeProject.AI Version 2.0

OK. If you are going to try 2.07 then you need to be on the latest version of BI. There is no need to add models or move them, you will just screw up the CPAI. To use the yolov5l you must use "Default object detection" Then just select what size of model you want. This is why we suggest to delete the folders when re installing CPAI to be sure everything is in the right place.

I am on latest and greatest BI version. Been on that everyday as it comes out. I forgot to mention it here but I did completely uninstall and nuke the folders and installed CPAI 2.0.7. It still crashes. I’m not getting anywhere near the performance as I used to with just throwing that yolov5l in the custom models folder.

I tried the whole default object detection but it literally crashes the entire computer where it’s not even accessible or shows anything on the monitor. Definitely doing something to the GPU to fully crash it. I have a P1000 GPU. Everything was originally installed using the cuda install script off CPs website. So that has Cuda 12.0 installed.

So for now trying the IPcam-general model to see how that works. Just turned off default object detection. Maybe the yolov5l just isn’t working with this anymore. But still don’t get why it won’t even show it on the list. It’s seems like it’s ignoring it on purpose.

Guess we wait and see how ipcam-general works now. It’s been crashing usually within 8-12hrs. So should know quickly.


Sent from my iPhone using Tapatalk
 
Any help guys?
Yes, I've tried both 'Object Detection (YOLOv5 .NET)' and 'Object Detection (YOLOv5 6.2)' and neither works for the custom models.
I've even completely uninstalled CodeProjectAI and deleted the folder in Program Files and reinstalled.
 
Last edited:
Ok so upon the THIRD time of completely nuking the CPAI installation from my machine and reinstalling it, I've finally got it to do some analysis on alerts.
However, now I'm seeing HORRIFIC response times in the ~7000ms+ range using the ipcam-combined and ipcam-animal models.
I was previously getting ~150ms using the same models.

1675316610969.png

It appears that I can only get CPAI to work by using Object Detection (YOLOv5 6.2).
I've tried running it with the 'gpu' enabled and disabled and there's no difference.

I'm running an i5-11500 with16GB RAM and no GPU.
Rebooted multiple times.

Any suggestions?
 
Last edited:
Ok so upon the THIRD time of completely nuking the CPAI installation from my machine and reinstalling it, I've finally got it to do some analysis on alerts.
However, now I'm seeing HORRIFIC response times in the ~7000ms+ range using the ipcam-combined and ipcam-animal models.
I was previously getting ~150ms using the same models.

View attachment 152797

It appears that I can only get CPAI to work by using Object Detection (YOLOv5 6.2).
I've tried running it with the 'gpu' enabled and disabled and there's no difference.

I'm running an i5-11500 with16GB RAM and no GPU.
Rebooted multiple times.

Any suggestions?
Do you only have the Object Detection (YOLOv5 6.2) enabled? If you have more then one Object Detection module enabled it might cause the issue you are see.
 
  • Like
Reactions: Cooltiger
Below are the locations of the custom models also the link to my GitHub Repo if you need the models

Object Detection (YOLOv5 6.2)

View attachment 152758

Object Detection (YOLOv5 .NET)

View attachment 152759
Mike would it be possible to include delivery to that directory, on every update that I do, I need to remember to copy that file over. Thanks

Sent from my SM-S906U using Tapatalk
 
Okay guys, I got 1.6.8 installed and still have issues with the License Plate reader. I am at a loss. Evidently there is something wrong on my machine. Thanks for all your help.

View attachment 152781
I may be wrong but I don't think there is anything wrong with your machine, I think the plate reader module works with the latest updates but I may be wrong. I will try to get with this version the plate running and let you know. Do you have bu recognize other objects? I now get, cars, person and etc.

Sent from my SM-S906U using Tapatalk
 
I may be wrong but I don't think there is anything wrong with your machine, I think the plate reader module works with the latest updates but I may be wrong. I will try to get with this version the plate running and let you know. Do you have bu recognize other objects? I now get, cars, person and etc.

Sent from my SM-S906U using Tapatalk
I have tried the latest updates and this update. I have uninstalled and re-installed numerous times. For some reason the ALPR module won't work on my machine. I will re-install again and see what it actually recognizes now. Originally in the first screen, it could find the plate and say Day Plate but no recognition. I then went to the image tab and it doesn't find anything. Somewhere up this thread I have posted the errors it posts.
 
Starting to see a repeating issue. I'm testing the latest version of BI and CPAI, and after running for several hours, it starts failing. Here's a screen shot of the error on the BI machine, log file right before the issue and Event Viewer BI error where it detects successfully and then starts crapping out. Any thoughts, is this similar to what some of you are experiencing right now?

Screen Shot 2023-02-02 at 6.17.58 PM.png

3 2/2/2023 5:47:36.208 PM Mailbox_Cam MOTION_A
0 2/2/2023 5:47:37.969 PM Mailbox_Cam AI: [ipcam-general] vehicle:92% [0,118 150,239] 1407ms
10 2/2/2023 5:47:38.075 PM Server [::]: Connected
0 2/2/2023 5:47:39.379 PM Mailbox_Cam Push: OK to My_iPhone
0 2/2/2023 5:48:47.935 PM New Move: 2 items 48.0M [210.0G/210.0G, 12.0G free]
0 2/2/2023 5:48:47.958 PM Alerts Move: nothing to do [10.0G/10.0G, 12.0G free] 2315 locked
0 2/2/2023 5:48:48.055 PM Stored Delete: 2 items 69.3M [220.0G/220.0G, 82.0G free]
0 2/2/2023 5:48:48.125 PM Alerts Delete: nothing to do [10.0G/10.0G, 12.0G free]
3 2/2/2023 5:53:34.374 PM Back_Yard_Cam MOTION_A
0 2/2/2023 5:53:37.431 PM Back_Yard_Cam AI: Alert cancelled [nothing found] 1428ms
0 2/2/2023 5:53:49.747 PM New Move: nothing to do [209.9G/210.0G, 12.0G free]
0 2/2/2023 5:53:49.780 PM Alerts Move: nothing to do [10.0G/10.0G, 12.0G free] 2315 locked
0 2/2/2023 5:53:49.781 PM Stored Delete: nothing to do [219.9G/220.0G, 82.1G free]
0 2/2/2023 5:53:49.859 PM Alerts Delete: nothing to do [10.0G/10.0G, 12.0G free]
3 2/2/2023 5:58:07.910 PM Mailbox_Cam MOTION_A
0 2/2/2023 5:58:09.989 PM Mailbox_Cam AI: Alert cancelled [nothing found] 475ms
0 2/2/2023 5:58:50.401 PM New Move: 1 items 22.2M [210.0G/210.0G, 12.0G free]
0 2/2/2023 5:58:50.424 PM Alerts Move: nothing to do [10.0G/10.0G, 12.0G free] 2315 locked
0 2/2/2023 5:58:50.515 PM Alerts Delete: nothing to do [10.0G/10.0G, 12.0G free]
0 2/2/2023 6:02:00.178 PM App Current profile: Sunset to 11 PM
0 2/2/2023 6:03:51.362 PM New Move: nothing to do [209.9G/210.0G, 12.0G free]
0 2/2/2023 6:03:51.384 PM Alerts Move: nothing to do [10.0G/10.0G, 12.0G free] 2315 locked
0 2/2/2023 6:03:51.459 PM Alerts Delete: nothing to do [10.0G/10.0G, 12.0G free]
3 2/2/2023 6:07:45.351 PM Mailbox_Cam MOTION_A
2 2/2/2023 6:07:45.357 PM Mailbox_Cam Clip: H264 encoder missing (848x480)
1 2/2/2023 6:07:55.310 PM Mailbox_Cam Alert image not created
2 2/2/2023 6:07:55.383 PM Mailbox_Cam Clip: H264 encoder missing (848x480)
0 2/2/2023 6:07:57.256 PM Mailbox_Cam AI: Alert cancelled [nothing found] 1888ms
0 2/2/2023 6:08:51.850 PM Alerts Move: nothing to do [10.0G/10.0G, 12.0G free] 2315 locked
0 2/2/2023 6:08:51.921 PM Alerts Delete: nothing to do [10.0G/10.0G, 12.0G free]
3 2/2/2023 6:11:28.547 PM Mailbox_Cam MOTION_A
2 2/2/2023 6:11:28.563 PM Mailbox_Cam Write to S:\Alerts\Mailbox_Cam.20230202_181128.0.3-1.jpg failed 2/1455
2 2/2/2023 6:11:28.749 PM Mailbox_Cam AI: not responding
1 2/2/2023 6:11:28.749 PM Mailbox_Cam AI: Alert cancelled [AI: not responding] 0ms
0 2/2/2023 6:11:58.031 PM App Nvidia video hardware detected
2 2/2/2023 6:11:58.035 PM App Restarted after unexpected shutdown (5.6.9.7/1500/240)
0 2/2/2023 6:12:01.016 PM App Current profile: Sunset to 11 PM
0 2/2/2023 6:12:17.042 PM New Move: 1 items 26.6M [210.0G/210.0G, 12.0G free]
0 2/2/2023 6:12:17.065 PM Alerts Move: nothing to do [10.0G/10.0G, 12.0G free] 2315 locked
0 2/2/2023 6:12:17.269 PM Stored Delete: 2 items 42.1M [220.0G/220.0G, 82.1G free]
0 2/2/2023 6:12:17.340 PM Alerts Delete: nothing to do [10.0G/10.0G, 12.0G free]
0 2/2/2023 6:12:17.341 PM Snapshots Delete: nothing to do [1.99G/2.00G, 2.00T free]
0 2/2/2023 6:12:17.341 PM Stored_Alerts Delete: nothing to do [16.5G/20.0G, 2.00T free]
4 2/2/2023 6:12:17.953 PM Clips S:: 21631, 219.9G/220.0G (+12.0G)
4 2/2/2023 6:12:17.954 PM Clips \\QNAP: 10291, 219.9G/220.0G (+82.1G)
4 2/2/2023 6:12:17.954 PM Clips C:: 584, 1.99G/2.00G
4 2/2/2023 6:12:17.954 PM Clips \\QNAP: 53126, 16.5G/20.0G
0 2/2/2023 6:12:20.104 PM App AI has been started
4 2/2/2023 6:12:20.654 PM BYR Signal: restored
0 2/2/2023 6:12:22.604 PM AI responding at 127.0.0.1:32168
4 2/2/2023 6:12:26.447 PM BYR Signal: restored
0 2/2/2023 6:12:27.131 PM App Today's sunrise: 7:37 AM, sunset: 6:02 PM
0 2/2/2023 6:12:27.145 PM Server Started on port 8347
3 2/2/2023 6:15:41.467 PM Front_Porch_Cam MOTION_A
2 2/2/2023 6:15:41.483 PM Front_Porch_Cam Write to S:\Alerts\Front_Porch_Cam.20230202_181541.0.3-1.jpg failed 2/1455
2 2/2/2023 6:15:41.669 PM Front_Porch_Cam AI: not responding
1 2/2/2023 6:15:41.670 PM Front_Porch_Cam AI: Alert cancelled [AI: not responding] 0ms
3 2/2/2023 6:15:46.089 PM Doorbell_Cam MOTION_A
2 2/2/2023 6:15:46.093 PM Doorbell_Cam Clip: H264 encoder missing (768x576)
1 2/2/2023 6:15:56.095 PM Doorbell_Cam Alert image not created
2 2/2/2023 6:15:56.121 PM Doorbell_Cam Clip: H264 encoder missing (768x576)
0 2/2/2023 6:15:56.729 PM Doorbell_Cam AI: Alert cancelled [nothing found] 525ms
Screen Shot 2023-02-02 at 6.28.23 PM.png
 
Starting to see a repeating issue. I'm testing the latest version of BI and CPAI, and after running for several hours, it starts failing. Here's a screen shot of the error on the BI machine, log file right before the issue and Event Viewer BI error where it detects successfully and then starts crapping out. Any thoughts, is this similar to what some of you are experiencing right now?

Ive seen the same issue. Constantly crashes on that. Im testing it now not using the default object detection (using ipcam-general instead now) and its been stable now for 20hrs where before it would barely every go past 8-12hrs. Mine, like yours, would show same error, multiple it would crash so hard that cant even remote in or get actual video on a monitor connected, had to cold start it each time.
 
Ive seen the same issue. Constantly crashes on that. Im testing it now not using the default object detection (using ipcam-general instead now) and its been stable now for 20hrs where before it would barely every go past 8-12hrs. Mine, like yours, would show same error, multiple it would crash so hard that cant even remote in or get actual video on a monitor connected, had to cold start it each time.
Thanks for the reply. Mine crashes like this without using default object detection. I'm going to try rolling back to CPAI v2.0.2 and see how it runs. That version used to be very stable for me.
 
Thanks for the reply. Mine crashes like this without using default object detection. I'm going to try rolling back to CPAI v2.0.2 and see how it runs. That version used to be very stable for me.
What model you using? Tonight Im going to test something by using a different model than ipcam-general. Going to try to make it use the yolo5vl which is what i was using before fine on 1.6.8.
 
Today just out of curiosity I installed Docker desktop and then installed the docker CPAI version. Wow, everything just worked fine without doing much reading. In case someone wants to give it a shot here was my process.

1. Go to: Get Docker | Docker Documentation and install Docker Desktop for Windows.
2. You will need to create a account ...no big deal.
3. Install Docker. Once it is done you will need to reboot. (also note, I used the WSL 2 method)
4. Skip the tutorial and just open a command prompt (admin) and past this:

docker pull codeproject/ai-server (this is CPU only mode version)

you will see it start the download process and let it do its thing.

5. Start the image within docker desktop and you will see a options window pop up. All I did was put in the port 32168 and hit start. Thats it, I tested it and it works just like the windows version.

Here is some more info as well:

Running CodeProject.AI Server in Docker - CodeProject.AI Server v2.0.6.

But this page says to do more than you need. I did not mess with anything other than tell it what port to use.

FWIW after I tried and tested Docker, I uninstalled all of it. I just don't see the need for it in my case as the windows version works great for me.
 
Last edited:
Today just out of curiosity I installed Docker desktop and then installed the docker CPAI version. Wow, everything just worked fine without doing much reading.

Does it chew less than literally all of your CPU resources every time it detects something?