sebastiantombs
Known around here
Just remember flagged alerts don't get deleted during regular disk cleanup. At 4% why bother, BI is detecting at that kind of level anyway.
Hell Yeah! Direct Deepstack Integration - 5.4.0 - March 31, 2021
- Thread starter fenderman
- Start date
Ssayer
BIT Beta Team
Like I said, for future reference. Got to have a benchmark... I don't have Auto-protect when flagged ticked, so the flagged alerts do get deleted.
Got this all set up and running the other day, very happy. But the alerts when I double click them to go to the alert time it takes me to the wrong time in the video feed. Example the alert is at 9:15, so I click the alert image and it takes me to 9:10 where there is no activity.
Anyone else have this problem or know of a fix?
My other request would be for mqtt options for the confirmed alerts. So fire a mqtt for a car or person if they are identified.
Anyone else have this problem or know of a fix?
My other request would be for mqtt options for the confirmed alerts. So fire a mqtt for a car or person if they are identified.
cvd
n3wb
I'm having a problem in the iOS app where any alerts which went through the Deepstack cannot be viewed in the app.
The alerts are fine in BI and the browser, but in the IOS app, clicking them shows the native iOS video browser with just a spinning icon.
If I turn off the Deepstack integration, the alerts are back to normal.
Has anyone else seen this?
BI 5.4.2.0. Latest IOS version as of now.
The alerts are fine in BI and the browser, but in the IOS app, clicking them shows the native iOS video browser with just a spinning icon.
If I turn off the Deepstack integration, the alerts are back to normal.
Has anyone else seen this?
BI 5.4.2.0. Latest IOS version as of now.
Hello everyone,
I am new to BI and IP cameras in general, I've started tinkering over a month ago and installed a few around my place. On my main profile, I have the outdoor cameras with motion detection enabled, but nothing to happen under the alert tab. In my case, I just want to have in the database highlights of when there was motion, without any other needs to enable alerts. I noticed after some troubleshooting, that deepstack doesn't work unless I change the "Fire when" to "this camera is triggered" under alerts. Which is fine, but is it an expected behavior? my purpose is to reduce false alarms and limit my triggers to people.
Problem number 2... Faces... This doesn't work unless I remove the substream configuration from BI for the specific camera. I guess the substream resolution is bad enough, that deepstack doesn't detect faces. And I cannot disable substreams or my CPU overloads when looking at all feeds. Is there a way to tell BI to use the mainstream for deepstack? all cams are hikvision 8mp running at 15fps full resolution, highest quality, 1/100 exposure (can't remember what the substreams are running on but it's the default value).
I am new to BI and IP cameras in general, I've started tinkering over a month ago and installed a few around my place. On my main profile, I have the outdoor cameras with motion detection enabled, but nothing to happen under the alert tab. In my case, I just want to have in the database highlights of when there was motion, without any other needs to enable alerts. I noticed after some troubleshooting, that deepstack doesn't work unless I change the "Fire when" to "this camera is triggered" under alerts. Which is fine, but is it an expected behavior? my purpose is to reduce false alarms and limit my triggers to people.
Problem number 2... Faces... This doesn't work unless I remove the substream configuration from BI for the specific camera. I guess the substream resolution is bad enough, that deepstack doesn't detect faces. And I cannot disable substreams or my CPU overloads when looking at all feeds. Is there a way to tell BI to use the mainstream for deepstack? all cams are hikvision 8mp running at 15fps full resolution, highest quality, 1/100 exposure (can't remember what the substreams are running on but it's the default value).
Regarding faces, it isn’t of any value now if you are using sub streams. The image captured is a pixelated mess. I have turned mine off.
Bluediamond
Young grasshopper
So it's been a bit since I wrote on the forums and I have been using Sentry for a super long time what are pro's and cons now is Deepstack better then Sentry or is it the beginning of a possible deprecation of Sentry from BI. Should I stop using sentry and make the switch to Deepstack?
Keep in mind that the same principles apply regarding pixels per foot (PPF) in order to get good faces. A 2.8mm lens trying to facial ID someone 45 feet away with this software is not going to work. If you have areas of coverage that are optically zoomed to make out faces, this is working very well, at least in my case. I have a few key pinch points that are working well at 256bitrate. As always, YMMV.
Keep in mind you can also up the substream quality a bit without serious CPU impacts. Try adjusting resolution and bitrates of the substream and see if it improves.
Many here have already gave up on Sentry based on this Deepstack integration. Many gave up on it even before Deepstack as Sentry seemed to be MIA...
kklee
Pulling my weight
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
That looks how it should for Object Detection, which is the API BI is using currently
where do you find the trigger settings for deep stack?
1. In General Settings -> AI tab
2. in Camera settings -> Trigger -> Artificial Intelligence
Only confirmed/cancelled by Deepstack. I would imagine filtering may come fairly soon, as it's a logical improvement to this new implementation.
Have you downloaded Deepstack already?
Unknown faces (if detected) will be stored in the folder you have selected in the settings.
TL1096r
IPCT Contributor
- Jan 28, 2017
- 1,211
- 469
oh.. i thought it was all built in with the new update... feeling stupid
You're not the only one, myself included
Using DeepStack with Windows 10 (CPU and GPU)
docs.deepstack.cc
sebastiantombs
Known around here
Being a stubborn old coot and not wanting to give up I started fooling, yet again, with DeepStack. I think I got it working properly, finally. In the interest of saving someone, anyone, the same frustrations this is what I ended up doing.
The "confidence" level is set at a default of 50%. I have it down at 5%, one tenth the original setting. Now it's detecting 100% of the valid triggers during the day, at least for me. I also set the additional views to one s that it actually "looks" at two frames one second apart. With matching iframe and frame rates and because Blue Iris starts recording on a full frame it gets two frames of full data to look at. Don't be afraid to lower that "confidence" setting to really low numbers is the bottom line.
This hasn't seemed to have cured the problem at night though. I just checked and it started missing clips about 20 minutes after sundown. It's still missing vehicles and it also missed a herd of deer that passed by one of the cameras. I think this is a contrast and scene lighting issue. In full daylight it works very well, at night in B&W not well at all. Given that it's a free app I think the daytime utility it has is more than worth the time and effort needed to set it up. I will continue playing with the "confidence" setting at night to see if I can get it to detect reliably.
The "confidence" level is set at a default of 50%. I have it down at 5%, one tenth the original setting. Now it's detecting 100% of the valid triggers during the day, at least for me. I also set the additional views to one s that it actually "looks" at two frames one second apart. With matching iframe and frame rates and because Blue Iris starts recording on a full frame it gets two frames of full data to look at. Don't be afraid to lower that "confidence" setting to really low numbers is the bottom line.
This hasn't seemed to have cured the problem at night though. I just checked and it started missing clips about 20 minutes after sundown. It's still missing vehicles and it also missed a herd of deer that passed by one of the cameras. I think this is a contrast and scene lighting issue. In full daylight it works very well, at night in B&W not well at all. Given that it's a free app I think the daytime utility it has is more than worth the time and effort needed to set it up. I will continue playing with the "confidence" setting at night to see if I can get it to detect reliably.
I created a folder (I used Aux 8 and renamed it) for unknown faces so I could keep them separate.