IP Cam Talk Custom Community DeepStack Model

[Corvus85]

For those of you who are running both the 'combined' model and the 'exdark' models, can you tell me if I've set this up correctly?

This is a screenshot taken from a camera set to my 'night' profile.
As I understand it, both the default 'objects' and 'combined' models should be inactive using these commands, leaving only the 'dark' model. Right?

 

[wittaj]

For those of you who are running both the 'combined' model and the 'exdark' models, can you tell me if I've set this up correctly?

This is a screenshot taken from a camera set to my 'night' profile.
As I understand it, both the default 'objects' and 'combined' models should be inactive using these commands, leaving only the 'dark' model. Right?

View attachment 117153

That would be correct.

 

[Corvus85]

That would be correct.

Interesting. So then why would tags from other models appear when selecting 'analyze with deepstack' when playing back a cancelled alert? Shouldn't it only be running the model that is active on the profile?

 

[wittaj]

Interesting. So then why would tags from other models appear when selecting 'analyze with deepstack' when playing back a cancelled alert? Shouldn't it only be running the model that is active on the profile?

Keep in mind that the "Analyze with Deepstack" under "Testing & Tuning" will ALWAYS perform better than live as it is after the fact and should not be used as an analysis tool to try to figure out why it didn't see and trigger for a car or person and will use all of the models you have. It should only be used to see what DeepStack can find in that clip, like "hmm I wonder if DeepStack can find a toothbrush" and then walk around with a toothbrush and have it identify it.

Don't believe me, run "Analyze with Deepstack" on a camera not using Deepstack and it will show EVERYTHING that Deepstack has in its objects to find that it sees in the clip. This method will show you EVERY ITEM deepstack searches for with all the models you have in the models folder.

It is the .DAT file that you need to be looking at that will show only the models being used on the cancelled alert for that camera.

 

[Corvus85]

Yeah I knew that, although I thought it was still limited to all the objects in the chosen model, not all models.

 

[MikeLud1]

If you are not using DeepStack default objects model in any of your cameras you can uncheck Default object detection. This will save on memory since Blue Iris will not load this model.

 

[Corvus85]

This is what I see when I CTRL+click on an alert.



Here's the settings for that cam.



Can someone tell me why Deepstack status says it's still running both 'dark' and 'combined' models? I've unchecked 'Default object detection' in AI settings.

 

[MikeLud1]

View attachment 117173
View attachment 117182

Can someone tell me why Deepstack status says it's still running both 'dark' and 'combined' models? I've unchecked 'Default object detection' in AI settings.

During playback Testing & Tuning DeepStack it will use all models.

 

[Corvus85]

During playback Testing & Tuning DeepStack it will use all models.

That doesn't make sense. This is literally the .dat file of what actually happened. Isn't this telling me that both 'dark' and 'combined' models were run? It even tells me how long they took to process.

 

[MikeLud1]

That doesn't make sense. This is literally the .dat file of what actually happened. Isn't this telling me that both 'dark' and 'combined' models were run? It even tells me how long they took to process.

Try holding down the CTRL key and double click on the alert, this will open up Deepstack status window for that alert. This should only show the models used for that alert.

 

[Corvus85]

Try holding down the CTRL key and double click on the alert, this will open up Deepstack status window for that alert. This should only show the models used for that alert.

But that's literally what I did to get that screenshot.

Any ideas?

 

[MikeLud1]

But that's literally what I did to get that screenshot.

Any ideas?

Something is not right the highlighted should end in dat not bvr. Are you doing the CTRL + double click on the alert

 

[MikeLud1]

But that's literally what I did to get that screenshot.

Any ideas?

Also make sure you have Testing & Tuning turned Off

 

[Corvus85]

Also make sure you have Testing & Tuning turned Off

Ok that seemed to do the trick. Thanks! Didn't realize I still had it enabled

 

[Lucas Tam]

HI all,

DeepStack is randomly failing on me after 4 or 5 days of uptime ... Blue Iris continues to function but when it tries to submit an image for AI processing, nothing is returned - I notice the processing time is very short, i.e 50ms - 60ms.

BI doesn't seem to log a critical error. The DeepStack Web Interface loads properly when I press the "Test" button in BI.

However, if I manually test a clip via playback in BI, I get the following error in the DeepStack BI Log Viewer:

[
{
"api":"general",
"found":{
"success":false,
"error":"error occured on the server",
"duration":0}
}
]

Any way to debug why DeepStack dies?

 

[jNewcs]

I've gone through this thread but was wondering if there is a straight forward walkthru on getting this setup? I've gotten bits and pieces but was hoping for a start to finish.
I like the idea of having AI identify more common animals and such.
Thanks

 

[CrazyAsYou]

I have no need to track anything other than people and cars so I'm using the general.pt (v1.2) - I'm using the GPU version of deepstack with an old GTX 970 graphics card, I send the substream images and the results are excellent, process times average ~70ms some are as low as 30ms.

 

[cbm214]

I'm in need of some beginners help using these custom files and how to set up BI, can someone help guide me?

What I've done:
1. Downloaded the three .pt file from Git (animal, combined and general.pt) and saved into a folder I created C:\Deepstack\MyModels\
2. In BI, general settings -> AI tab. I ticked the option 'Use custom model folder' and added the path to C:\Deepstack\MyModels
3. I left the option 'Default object detection' ticked, I'm not sure this needs unticking as to me ticking this option means 'use the .pt files that ship with Deepstack, the ones that include objects like bananas, broccoli', the ones we don't want to use now as we want to use the smaller list of objects in the three .pt downloaded from Git.
4. Restart the PC running BI and Deepstack.


I then tested everything it by walking out in front of some cameras. Looking in the BI log the Deepstack times were now longer, around 1.4 seconds instead of 500ms. I must of got something wrong.

Questions.
1. Should the 'Default object detection' tick box be ticked or not and why?
2. I've seen in some posts in this thread that you also need to edit each camera you wish to use these new .pt files for and add to the 'Custom Models:' field text like 'objects:0,combined'. Do you need to add this to the cameras or not?
3. If the answer to Q2 is yes, then what does it mean (yes I have read the BI manual and I understand adding :0 to the end of a text string excludes that model eg 'objects' is disabled), just what is 'objects', I'm assuming it's nothing to do with the tickbox option 'Default object detection' as that does use the word object but not objects? It's not the name of a .pt file, what does 'objects' mean?
4. I've also seen in other posts in this thread people have added 'dark' to the list in 'Custom Models:' field, again, where has this word come from, I don't see a file called dark.pt.
5. Log file, is there a different log for Deepstack analysis than the one in BI?

I feel I'm missing something simple and I have looked through all the 19 pages of this post for a 'How-to for dummies' guide but not noticed one. I would appreciate some step by step guide, others may appreciate one to.

Thanks in advance

 

[wittaj]

I'm in need of some beginners help using these custom files and how to set up BI, can someone help guide me?......

Questions. Answers in red
1. Should the 'Default object detection' tick box be ticked or not and why? If you are using the custom models and have no need for the exhaustive items in DS, then yes uncheck that box. Your return times will be much faster.
2. I've seen in some posts in this thread that you also need to edit each camera you wish to use these new .pt files for and add to the 'Custom Models:' field text like 'objects:0,combined'. Do you need to add this to the cameras or not? Yes you do. The reason your response times have increased is because you are now running the default model and all the custom models at the same time for each camera. Go into each camera and use just the model you need for that camera. The custom models you downloaded took out all the junk we don't need like broccoli, toothbrush, etc. LOL, so it allows for a quicker response. Using the default results in the image checking for EVERY item in that list, thus the custom model was created to speed up response times. It is counterproductive to use the both the default objects and the custom models you pulled at the same time.
3. If the answer to Q2 is yes, then what does it mean (yes I have read the BI manual and I understand adding :0 to the end of a text string excludes that model eg 'objects' is disabled), just what is 'objects', I'm assuming it's nothing to do with the tickbox option 'Default object detection' as that does use the word object but not objects? It's not the name of a .pt file, what does 'objects' mean? Objects is the default object detection model in DeepStack. If you do not want to use the default for any camera, then uncheck the box from question #1. If you want to use it for some cameras, then objects:0 has to go in the cameras you do not want the default one to use.
4. I've also seen in other posts in this thread people have added 'dark' to the list in 'Custom Models:' field, again, where has this word come from, I don't see a file called dark.pt. dark.pt is yet another model available in a different location than the custom models you pulled. The ones you pulled were developed by a member here. The dark.pt was developed by another person. GitHub - OlafenwaMoses/DeepStack_ExDark: A DeepStack custom model for detecting common objects in dark/night images and videos.
5. Log file, is there a different log for Deepstack analysis than the one in BI? No, they are all in the same log file. But if in each camera you check the save .DAT file option, then in BI you can pull up the DS specifics for that triggered analysis.

 

[cbm214]

Questions. Answers in red
1. Should the 'Default object detection' tick box be ticked or not and why? If you are using the custom models and have no need for the exhaustive items in DS, then yes uncheck that box. Your return times will be much faster.
2. I've seen in some posts in this thread that you also need to edit each camera you wish to use these new .pt files for and add to the 'Custom Models:' field text like 'objects:0,combined'. Do you need to add this to the cameras or not? Yes you do. The reason your response times have increased is because you are now running the default model and all the custom models at the same time for each camera. Go into each camera and use just the model you need for that camera. The custom models you downloaded took out all the junk we don't need like broccoli, toothbrush, etc. LOL, so it allows for a quicker response. Using the default results in the image checking for EVERY item in that list, thus the custom model was created to speed up response times. It is counterproductive to use the both the default objects and the custom models you pulled at the same time.
3. If the answer to Q2 is yes, then what does it mean (yes I have read the BI manual and I understand adding :0 to the end of a text string excludes that model eg 'objects' is disabled), just what is 'objects', I'm assuming it's nothing to do with the tickbox option 'Default object detection' as that does use the word object but not objects? It's not the name of a .pt file, what does 'objects' mean? Objects is the default object detection model in DeepStack. If you do not want to use the default for any camera, then uncheck the box from question #1. If you want to use it for some cameras, then objects:0 has to go in the cameras you do not want the default one to use.
4. I've also seen in other posts in this thread people have added 'dark' to the list in 'Custom Models:' field, again, where has this word come from, I don't see a file called dark.pt. dark.pt is yet another model available in a different location than the custom models you pulled. The ones you pulled were developed by a member here. The dark.pt was developed by another person. GitHub - OlafenwaMoses/DeepStack_ExDark: A DeepStack custom model for detecting common objects in dark/night images and videos.
5. Log file, is there a different log for Deepstack analysis than the one in BI? No, they are all in the same log file.

Wow, that was quick, thank you, that's been really helpful, I now understand things a lot better and will configure BI appropriately.

One question I forgot to add to my first post. In each camera there is the field 'Mark as vehicle', what does this mean, BI Help doesn't mention it? I've seen in this forum someone say it's if you use number plate recognition, what Deepstack items are classed a 'vehicle' and therefore get set off to be plate recognised, is this true? I've seen this field filled in in the thread, not sure what it's for.

Thanks once again for your help, I appreciate it.