Hell Yeah! Direct Deepstack Integration - 5.4.0 - March 31, 2021

Ssayer said:
Man! Something is very wrong with the facial recognition with this thing. I tested it walking all over hitting different cameras and... I mean, C'mon man! Perusing the images that it's picked out, all I see is one UGLY looking guy! :rofl: :rofl: :rofl:
Click to expand...
Is there a folder it's saving face recognitions to other than the one we manually save to?
 
  • Like
Reactions: Ssayer and fenderman
I don't see how (yet) or I would have already done so.
 
I'm presuming that you've gone out and tried it since selecting that folder? It's not like it's going to go back and move older files into it. If so, then I don't know. It works here (on both of the latest updates).
 
Ssayer said:
I'm presuming that you've gone out and tried it since selecting that folder? It's not like it's going to go back and move older files into it. If so, then I don't know. It works here (on both of the latest updates).
Click to expand...
Yea I'll keep trying
 
One other "gottcha", have you put "faces" in your notifications list for the camera AI?
 
I believe so. Otherwise the clip won't get tagged and recorded. The other guys using face detection are probably better sources for this and whether it's actually "face" or "faces", but you could give it a try in the mean time. Can't hurt anything.
 
sebastiantombs said:
The main stream for AI would help a lot I suspect. Varying the time from 1-5 seconds is nice as long as it's not whole numbers and allows tenths of a second. I'd also guess from an integration standpoint that it will be a little harder to add the main stream feature.
Click to expand...

I am using AiTool with DeepStack and I have a few suggestions that worked for me to improve object identification in DeepStack. First some information about my hardware. The BI PC is an I7-6700 with 32GB RAM. My cameras are IPC-T2431T-AS 3.6mm from EmpireTech, 4 in total. Below are my camera settings.

1617909675332.png

I am running 5 instances of the CPU version of DeepStack in a Docker container. Windows is configured for WSL2 for the Docker container. DeepStack MODE=High with processing times below.

1617911718376.png

Below is a typical resource usage without any motion. Docker is the VMMEM process and it is using most of the resources. The jpegs are sent to a 3GB RAM Disk and that is part of the 37% memory usage below. With all 4 cameras triggered the CPU usage gets up to 70%.

1617910326056.png
I have BI configured using both the main and sub stream. A cloned camera is detecting motion and generating the jpeg. I trigger the Clone Master with my Home Automation via url for event recording. The clone is using the sub stream to detect motion but generating a jpeg with the main stream. You accomplish jpeg generation using the main stream by enabling "Pre-trigger video buffer" I set the buffer to 2 seconds but I could lower that closer to 1 second. The jpegs are generated every 3 seconds with a break time of 6 seconds under the trigger settings so 3 images are taken and sent to DeepStack for each event. During daylight I have the image quality of the jpeg set to 50%. My minimum confidence is set at 42% for a person. In the daytime persons are identified with 70%-80% confidence. Occasionally my dog will be identified as a person in the daytime with a confidence as high as 40% which is why the confidence level for the person is at 42%.

1617911067374.png

After dark the BI profile changes and I boost the jpeg quality to 100%. The images at night are black and white so the jpeg size is actually smaller and processes about the same at night. The Home Automation system is set differently at night in that a push notification and text will not be sent out unless a person has been identified twice by any combination of cameras within 5 minutes. This is because DeepStack will identify the dog as a person with a high confidence level at night but very rarely twice in the 5 minute interval. Using the main stream for jpeg generation, increasing the image quality at night and running DeepStack in MODE=High improved my results.
 
  • Like
Reactions: beandip and shannondalebreaux
kklee said:
I've switched over to Deepstack from Sentry. Sentry tended to give a lot of false alerts since you can't tune anything. Deepstack is tuneable, and I expect that the BI developer will continue to improve the integration and tuning abilities. I won't be renewing my Sentry sub, which is expiring in May, so perfect timing. I'd rather send the funds towards BI to keep adding new features like this.

Give it a try! It's super easy, install Deepstack (no need to configure anything) and turn it on in BI. You may need to do some BI tuning, but if you're already using Sentry, trigger tuning should be minimal.

Since some people have been having problems finding the installation package, Here's the link: Release DeepStack Windows February Release · johnolafenwa/DeepStack
Click to expand...

I think I am going to wait a bit Sentry was truly easy just apply and go and It has not been to bad on the false positives. I would love to start tinkering with an AI but I think I am going to wait till BI starts to fine tune its properties.

And doing all this reading you need like two other tools to manage it properly.
 
  • Like
Reactions: looney2ns
@jz3082 First problem for me is five instances running in a docker. I don't want a docker and won't run one. Second problem is that you're analyzing jpg's, being captured separately, from what I see. It, apparently, is already getting snapshots when running in a Windows environment through BI.

Again, it's a nice little addition, but it is not totally reliable yet. Even to get some reliability takes some serious time and tinkering. For the "average" homeowner with a surveillance system it really isn't all that effective. Just my opinion based on the anecdotal results I see with it. YMMV.
 
  • Like
Reactions: beandip and looney2ns
Bluediamond said:
I think I am going to wait a bit Sentry was truly easy just apply and go and It has not been to bad on the false positives. I would love to start tinkering with an AI but I think I am going to wait till BI starts to fine tune its properties.

And doing all this reading you need like two other tools to manage it properly.
Click to expand...
In my usage case, it was a simple swap and I've found Deepstack to be better with regards to false alerts. No tuning required for what I want out of it. The hardest part was going into each camera config to change the AI setting to Deepstack instead of Sentry.
 
What's the trick to get it to not recognize static objects each time? It keep labeling my parked car, neighbors etc anytime there is movement.

Ideally of course it just highlights the new object to the scene. I realize this is exists, I'm just missing it.

Thank you!
 
I was looking around the DeepStack forum today and found a thread regarding the UI in Windows. This is the one that says it's an applet and it's in the readme file. I don't see anything in that readme file.

forum.deepstack.cc

Can deepstack-ui be run without docker?

Running windows 10 no docker on my remote pc ui hopeful.
forum.deepstack.cc forum.deepstack.cc
 
  • Like
Reactions: tech101
You must log in or register to reply here.
Top Bottom