I've had it up and running for a long while, stopped having reliable AI when versions 2.8 and 2.9.5 came out. Running an older Windows 10 PC with 30 cameras and was using the single coral TPU setup to off load the AI overhead with no problems prior. No matter what I've tried I cant get it to work reliably with 2.9.5. Multiple issues randomly pop up and stop the coral module. I've clean installed multiple times, nothing gets it fixed. Is there any future in Code Project AI or is it done?
Is CodeProject.AI done?
- Thread starter nashvegas
- Start date
I understand from some posts that back in September / October 2024, Chris Maunder and Matthew Dennis are no longer with the Code Project AI Server project. The CodeProject.AI-Server Discussions · codeproject CodeProject.AI-Server · Discussion #132
Since that timeframe it has become stagnant, last release was in December of 2024 which came with a good amount of bugs with the 2.8 and 2.9.5 versions, some of which broke coral processing for users like me that prior worked just fine.
So based on the current trajectory compared to the past, I'm guessing its done, or on the back burner so far its just not going to get the attention to keep things moving forward.
I'm getting ready to build a windows 11 machine to move forward with, prior I was planning on putting hardware in to offload the AI work to a TPU or GPU for that machine, but at this point with the lack of support, I'm leaning toward this thing is done and what's the point?
Since that timeframe it has become stagnant, last release was in December of 2024 which came with a good amount of bugs with the 2.8 and 2.9.5 versions, some of which broke coral processing for users like me that prior worked just fine.
So based on the current trajectory compared to the past, I'm guessing its done, or on the back burner so far its just not going to get the attention to keep things moving forward.
I'm getting ready to build a windows 11 machine to move forward with, prior I was planning on putting hardware in to offload the AI work to a TPU or GPU for that machine, but at this point with the lack of support, I'm leaning toward this thing is done and what's the point?
That comes down to you and your tolerance/acceptance of using it.
If not CodeProject, then what do you use? Wait for BI to decide CodeProject is dead and moves to the next AI, whatever that is?
Many here still use Deepstack and it hasn't been touched in years. Actually that is why BI moved to CodeProject.
Many here believe in the don't fix what ain't broke philosophy so they don't touch it once it works and is stable until something is added they want or need.
If you had a system that was stable and working for you, then why update? Use the version that you said was working fine.
Or do as many of us have and slowly update to cameras with AI.
Unless you have special use cases or cameras without built-in AI and need to use CodeProject or Deepstack, you may find the camera AI to be better.
Many here use strictly the camera AI and get no false triggers and none of the CPU maxing out during rain events and don't wake up to hundreds of "no object find" triggers either.
You may find that the camera AI has got so good that doing CodeProject is kinda overkill and adds more complexity, time delay, and potential for issues.
Whether to use camera AI or BI AI is obviously up to you, but of course, the AI in the camera may be more than sufficient for your needs without needing BI AI. Do you need the orange box around every object? Do you want to identify animals or logos? Or is just human or vehicle sufficient.
The camera AI is useful to many people, but BI has way more motion setting granularity than the cameras, and some people need that additional detail, especially if wanting AI for more than a car or person. For folks that want AI and alerts on animals or specifically a UPS truck then they need the additional AI.
There isn't really a best practice because every field of view is different and use case and needs are different.
To many here, BI motion without AI is more than adequate for what they do.
To many here, camera AI is more than adequate for what they do.
To many here, using the BI AI adds additional functionality that the above alone can not do.
It comes down to testing with each field of view and which one gives you the most consistent results.
While some of that third party stuff is cool like tagging was it a dog or a bear, I don't need all that fancy stuff. If my camera triggers BI to tag an alert for human or vehicle and BI can accomplish what I need by way of a text or email or push or whatever, that is sufficient for my needs. I just want to be alerted if a person or vehicle is on my property and the camera AI does a fine job with that.
However, I do run BI AI on a few cameras so that it knocks out headlight shine so that the alert image includes the vehicle. The camera AI will trigger for a car, but the alert image was always just the headlights. I also run the ANPR AI module. I am running it with the CPU and not a GPU.
The true test....I have found the AI of the cameras to work even in a freakin blizzard....imagine how much the CPU/GPU would be maxing out sending all the snow pictures for analysis to CodeProject LOL. My non-AI cams in BI were triggering all night. This picture was ran through AI (without the IVS or red lines on it) and it failed to recognize a person in the picture, but the camera AI did. This pic says it all and the video had the red box over it even in complete white out on the screen:
See this thread on how using just Dahua AI may be sufficient for your needs (and other cameras with AI would perform similar):
Who uses Dahua AI capable cameras? Reliable AI for triggering events? Pro's/con's?
If not CodeProject, then what do you use? Wait for BI to decide CodeProject is dead and moves to the next AI, whatever that is?
Many here still use Deepstack and it hasn't been touched in years. Actually that is why BI moved to CodeProject.
Many here believe in the don't fix what ain't broke philosophy so they don't touch it once it works and is stable until something is added they want or need.
If you had a system that was stable and working for you, then why update? Use the version that you said was working fine.
Or do as many of us have and slowly update to cameras with AI.
Unless you have special use cases or cameras without built-in AI and need to use CodeProject or Deepstack, you may find the camera AI to be better.
Many here use strictly the camera AI and get no false triggers and none of the CPU maxing out during rain events and don't wake up to hundreds of "no object find" triggers either.
You may find that the camera AI has got so good that doing CodeProject is kinda overkill and adds more complexity, time delay, and potential for issues.
Whether to use camera AI or BI AI is obviously up to you, but of course, the AI in the camera may be more than sufficient for your needs without needing BI AI. Do you need the orange box around every object? Do you want to identify animals or logos? Or is just human or vehicle sufficient.
The camera AI is useful to many people, but BI has way more motion setting granularity than the cameras, and some people need that additional detail, especially if wanting AI for more than a car or person. For folks that want AI and alerts on animals or specifically a UPS truck then they need the additional AI.
There isn't really a best practice because every field of view is different and use case and needs are different.
To many here, BI motion without AI is more than adequate for what they do.
To many here, camera AI is more than adequate for what they do.
To many here, using the BI AI adds additional functionality that the above alone can not do.
It comes down to testing with each field of view and which one gives you the most consistent results.
While some of that third party stuff is cool like tagging was it a dog or a bear, I don't need all that fancy stuff. If my camera triggers BI to tag an alert for human or vehicle and BI can accomplish what I need by way of a text or email or push or whatever, that is sufficient for my needs. I just want to be alerted if a person or vehicle is on my property and the camera AI does a fine job with that.
However, I do run BI AI on a few cameras so that it knocks out headlight shine so that the alert image includes the vehicle. The camera AI will trigger for a car, but the alert image was always just the headlights. I also run the ANPR AI module. I am running it with the CPU and not a GPU.
The true test....I have found the AI of the cameras to work even in a freakin blizzard....imagine how much the CPU/GPU would be maxing out sending all the snow pictures for analysis to CodeProject LOL. My non-AI cams in BI were triggering all night. This picture was ran through AI (without the IVS or red lines on it) and it failed to recognize a person in the picture, but the camera AI did. This pic says it all and the video had the red box over it even in complete white out on the screen:
See this thread on how using just Dahua AI may be sufficient for your needs (and other cameras with AI would perform similar):
Who uses Dahua AI capable cameras? Reliable AI for triggering events? Pro's/con's?
The BI and Code Project AI marriage was great (while it lasted I suppose now), both was putting efforts into frequent advances via the UI or new and improved models / capabilities. I'm not sure what the answer will be for BI AI moving forward. If AI is going to be a priority for BI, then someone needs to pay the piper and not rely on volunteer work, people get burnt out and move on, such is the story of life. That's not going to be much of a sustainable model. I'm just going to consider that this marriage is dead (until proven otherwise), and what we got in the current version is all we are likely going to have until BI figures out what to do with its future with AI.
Sybertiger
Known around here
Wow, this is news to me. I haven't been following CPAI as I'm still using DS. My plan was eventually to upgrade my BI computer to a newer Win11 machine when BI 6.0 came out and move to CPAI as I got the impression that it was better than DS. How disappointing. I guess it'll be the wild west for some time to come with AI solutions.
Code Project is like an old dinosaur compared to other on prem solutions available today. The world of AI has exploded with quantum leaps in just the last couple of years yet Code Project is virtually unchanged. I'm still using Beta 2.0.8 and just looked at the release notes to see if there were any major features or improvements that would make it worth upgrading and instead find that 2.8+ is considered more buggy and unreliable let alone having more accurate recognition of people and things.
I'd love to see BI have with a more generalized plugin API to expand third party integrations.
I'd love to see BI have with a more generalized plugin API to expand third party integrations.
Sybertiger
Known around here
I think "free" is the key. Most of us are not interested in a monthly subscription service. I'll change out all my old cams and go with new cams with built-in AI before I pay for a subscription.
That's why I said on-prem which usually doesn't have a subscription. Rekognition is still reasonably priced with 5000 images for free in tier and then $0.001 per image for the first million images in tier 1. For my purposes, the free tier is fine since I only hit 3 to 4k motion events per month. Even if there wasn't a free tier, 5K images would be $5 / month.
Still I'm allergic to subscriptions. I only do 2 streaming services at a time and cycle through them. I don't have to see the latest season of my favorite shows on each as they become available.
MikeLud1
IPCT Contributor
The last time I chatted with Chris from CP.AI was around the holidays, he did say that him and Matthew are still going to keep developing CP.AI. Chris has recently closed some issue in the GitHub repo.
Last edited:
Any update in when support for a YOLO11 module will be added?
Also, I can't seem to get any reply on when (if?) my delivery module will ever be incorporated in the preloaded custom models.
GitHub - VideoDad/AI-Models: Custom Trained AI Models for object detection
Custom Trained AI Models for object detection. Contribute to VideoDad/AI-Models development by creating an account on GitHub.
github.com
All my requests in this area have gone unanswered.