CodeProject.AI Version 2.0

Status update: Uninstall/reinstall took 65 min. ALPR just finished installing and now shows Started. That took another 16 min.

Dave Lonsdale said:
I have the same problem - a long copper overhead line. Openreach intended to upgrade it to fibre over 18 months ago for free but when they hit their 80% target in Congleton they moved on. Now it’s 2025. Are the ISP’s offering to install fibre for you for a price? If so, what is the price? Starlink would be too expensive for me.
Click to expand...
One ISP quotes their install price as $75 and 1 Gbps for $90/month or 300 Mbps for $70.
The other one either doesn't charge for install, or aren't saying. They offer 2 Gbps for $100/month or 1 Gbps for $80/month.
 
  • Like
Reactions: gwminor48
Any help on this?

Setting up CPAI for use with my BI install. BI is run in a VM on Host1 and plan to run CPAI on another machine Host2 on the network (cpai-host-ip : port 32168) with GPU/CUDA enabled on an Nvidia 3060 GPU.
My question is about how to set this up properly ... Configuring the location of AI is clear in the BI setting panel (ip : port 32168) but how are you supposed to deal with the models known to BI when CPAI is remote.... I know the CPAI install will automatically download and populate the models locally on Host2. How will BI on Host1 have any idea of what is installed?

I had read

So: Blue Iris doesn't currently change CodeProject.AI Server's settings, so it doesn't provide you a way to change the custom model folder location from within Blue Iris.

Blue Iris will still use the contents of this folder to determine the calls it makes. If you don't specify a model to use in the Custom Models textbox, then Blue Iris will use all models in the custom models folder that it knows about.
Click to expand...
Can someone please provide me the the proper steps/guidance on how to configure the remote CPAI instance properly in BI?

Thanks in advance
 
Last edited:
Stephen Flamm said:
Any help on this?

Setting up CPAI for use with my BI install. BI is run in a VM on Host1 and plan to run CPAI on another machine Host2 on the network (cpai-host-ip : port 32168) with GPU/CUDA enabled on an Nvidia 3060 GPU.
My question is about how to set this up properly ... Configuring the location of AI is clear in the BI setting panel (ip : port 32168) but how are you supposed to deal with the models known to BI when CPAI is remote.... I know the CPAI install will automatically download and populate the models locally on Host2. How will BI on Host1 have any idea of what is installed?

I had read


Can someone please provide me the the proper steps/guidance on how to configure the remote CPAI instance properly in BI?

Thanks in advance
Click to expand...
To get the custom models to work. extract the attached to folder "custom models" on Host1 then point Blue Iris to use this folder for custom models.

I recommend emailing Blue Iris support and explain the issues you are have with using CP.AI on a remotes Host. He should be able to fix it, someone else is having a similar issue using CP.AI in a Docker container and I asked them to email BI support so if Ken see enough of this issue it will help get it addressed sooner.
 

Attachments

Last edited:
  • Like
Reactions: looney2ns, tech101 and fenderman
harleyl7 said:
Once we have it installed, how do we use the LPR? Also, does it log license plates?
Click to expand...
See the below post. The plate will be logged in BI log

1674702333147.png

CodeProject.AI Version 2.0

CodeProject.AI version 2.0 was released Jan 16, 2023, this thread is for all topic CodeProject.AI version 2.0. https://www.codeproject.com/Articles/5322557/CodeProject-AI-Server-AI-the-easy-way If you need any info on CodeProject.AI version 1.x please refer to the below link (there is some...
ipcamtalk.com ipcamtalk.com
 
  • Like
Reactions: looney2ns, tech101, samplenhold and 1 other person
MikeLud1 said:
To get the custom models to work. extract the attached to folder "custom models" on Host1 then point Blue Iris to use this folder for custom models.

I recommend emailing Blue Iris support and explain the issues you are have with using CP.AI on a remotes Host. He should be able to fix it, someone else is having a similar issue using CP.AI in a Docker container and I asked them to email BI support so if Ken see enough of this issue it will help get it addressed sooner.
Click to expand...

Thanks for the help and guidance.

I will send this over to Blue Iris support - is the best way to inform them to email support@blueirissoftware.com or is there another address specifically for Ken?
 
If I may, I would like to compliment everyone here and the team at codeproject, great job. The dedication to what they do is impressive.
Today, I even got my plate day plate to be recognized. How do I get to recognize the numbers? Do I need to install the OCR module?

Thanks

Sent from my SM-S906U using Tapatalk
 
  • Like
Reactions: looney2ns, JNDATHP and fenderman
Alan_F said:
Tried:
objects:0,actionnetv2:0,ipcam-animal:0,ipcam-combined:0,ipcam-dark:0,ipcam-general:0

Check my work... that is a direct cut and paste from the UI.

Results in all models being used and 'nothing found' result even though the analysis window shows listed 'to confirm' objects from the non-excluded custom models (in this case, there is a DayPlate in the image, and that string above excludes everything BUT license-plate)
Click to expand...

Replying to myself since I think I've figured part of this out now. and wanted to clear it up for anyone reading this later:

If you Ctrl-Double Click an alert, it opens the AI analysis window. If "Testing and Tuning - Analyze with AI" is on, the analysis window shows ALL the models and detects everything. If I turn off Testing and Tuning, the analysis window only shows the models and objects that were originally used to analyze the clip.

Knowing that, I was able to go back and test what was happening with the exclusion list (objects:0,actionnetv2:0,ipcam-animal:0,ipcam-combined:0,ipcam-dark:0,ipcam-general:0). When I use that list, it doesn't use any models. If I'm reading the docs right, it should have used license-plate, because anything with a '0' is an exclusion list:

By default, all Custom model files are considered, although you may select a subset here by
specifying one or more (comma delimited; omit the file extensions). Custom models are
enabled on the global Settings AI page. This list can be either an inclusion list or an exclusion
list That is, if any model is given the :0 designation, then all models are used except those
marked :0; however if there are no :0 entries, then only the models listed will be used. You
may also add faces:0 and/or objects:0 here to specifically skip face and/or default object
detection, but note that this then makes the list an exclusion list for the purpose of custom
models. Specify a single non-existent model name (“none”) and no custom models will be
used.
Click to expand...

So either I don't understand the docs, or it isn't quite working as written.
 
  • Like
Reactions: mk1shitrocket
Alan_F said:
Replying to myself since I think I've figured part of this out now. and wanted to clear it up for anyone reading this later:

If you Ctrl-Double Click an alert, it opens the AI analysis window. If "Testing and Tuning - Analyze with AI" is on, the analysis window shows ALL the models and detects everything. If I turn off Testing and Tuning, the analysis window only shows the models and objects that were originally used to analyze the clip.

Knowing that, I was able to go back and test what was happening with the exclusion list (objects:0,actionnetv2:0,ipcam-animal:0,ipcam-combined:0,ipcam-dark:0,ipcam-general:0). When I use that list, it doesn't use any models. If I'm reading the docs right, it should have used license-plate, because anything with a '0' is an exclusion list:



So either I don't understand the docs, or it isn't quite working as written.
Click to expand...
This works as intended for me but I still call out the custom model I DO want to use. The way I understand it works is that as soon as you call out a single custom model to exclude, you need to exclude all the models you do not want to use but when you do exclude models, you still need to include the one(s) you want to use.
 
  • Wow
Reactions: Nunofya
@MikeLud1, I think I saw somewhere you mention there was a plan to have the ipcam-general model includes the dark model as well. Is the ipcam-general model included in this latest version a combined version that also include dark models as well or is that still in the work?
 
tried installing code project and avast is blocking pip.exe from running as it is not on a safe exe list
 
  • Like
Reactions: fenderman and Coldair
I'm on BI 5.6.9.2 with CodeProjectAI 2.0.60, and all by cameras use ipcam-combined. Things seem to be working well enough, so thanks to all involved.
In Settings>AI to the right of the ticked box Use custom models: I have actionnettv2ipcam-animali I cannot see what else is in the box, there is an i at the end of what is visible, there are no commas or spaces.

Can someone provide more information about what should be in this box and how to check?

What is actionnetv2, an available model that can be installed, presumably it already is. What are its properties?

I don't need license plate recognition however, what Installed Modules in the AI Dashboard would be good to have installed?

In a camera's AI what is recommended in the box To cancel, banana or Nothing found:0 ?

What would be reasonable for a camera's AI setting for the number of real time images to be processed and time to analyze one each? My T600 GPU manages in Status after 11682 Analyses a time of 135ms, that's 60% slower than before the V2 CodePoject AI on my same settings.
 
What's the best settings? Can anyone share that here. I'm mainly talking about in the camera AI section for what I need to place. For reference I'm running an i5-6500 with the igpu

Also, does my lpr camera have to be recording to work with lpr?
 
Last edited:
You must log in or register to reply here.
Top Bottom