IP Cam Talk Custom Community DeepStack Model

I just posted combined V2.0 in the first post. I am not sure how well the new animals will work due to the low image count. Please let me know how V2.0 works for you.
MikeLud1 said:
I figured out how to download the images that DeepStack used to train the object model. With these images I can start to make the community custom DeepStack model, below are the steps to create the custom model.

The first step would be to create a new DeepStack custom model using the same image that DeepStack used but removing all the labels we do not want. Below is a list of the labels that I think we should keep and a list of labels we should remove. Please let me know if I should change the lists, I am planning to create the new model this weekend so let me know by Friday if I should change the list.

The second step would be to start adding new labels with images the everyone contributes to the custom model.

Combined LabelsGeneral LabelsAnimal LabelsLabels removed from Deepstack original model
person, bicycle, car, motorcycle, bus, truck, bird, cat, dog, horse, sheep, cow, bear, deer, rabbit, raccoon, fox, skunk, squirrel, pig

Future Labels:
coyote, possum		person, bicycle, car, motorcycle, bus, truckbird, cat, dog, horse, sheep, cow, bear


Future Labels:
deer, rabbit, raccoon, fox, coyote, possum, skunk, squirrel, pig		airplane, train, boat, traffic light, fire hydrant, stop sign, parking meter, bench, elephant, zebra, giraffe, backpack, umbrella, handbag, tie, suitcase, frisbee, skis, snowboard, sports ball, kite, baseball bat, baseball glove, skateboard, surfboard, tennis racket, bottle, wine glass, cup, fork, knife, spoon, bowl, banana, apple, sandwich, orange, broccoli, carrot, hot dog, pizza, donut, cake, chair, couch, potted plant, bed, dining table, toilet, tv, laptop, mouse, remote, keyboard, cell phone, microwave, oven, toaster, sink, refrigerator, book, clock, vase, scissors, teddy bear, hair dryer, toothbrush
​
Combined V2.0 Training Results:​
Combined V1.1 Training Results:​
​
General V1.1 Training Results:​
Click to expand...
 
  • Like
Reactions: Kensko, 105437, BORIStheBLADE and 1 other person
If it is a good as 1.x versions then I am excited! The first ones were excellent. I used to get a few "person" flags of people in cars driving by my house on the stock/dark objects model. On yours, I got several alone today that I noticed. I had to look twice at what DS was saying was a person in the video. Sure enough it was catching the passengers in the vehicles. I am using these instead of Dark at night now. Dark seems to get higher % per object, but I get missed detections. The dreaded "nothing found" and "occupied" when there was something to see/catch. This would happen on both dark and stock models when used separately. I was having to use both stock + dark (as a few other also do) to catch almost all the cars driving by at night. (due to headlights, bad angles, etc) After I retuned my motion detection to work with your combined model I did not have a single miss last night using just combined 1.1. Good work and thank you for all your time and efforts to take our DeepStack detections to the next level!!
 
  • Like
Reactions: BadKarma, sebastiantombs and MikeLud1
Being a creature of habit I always use "objects:0,combined" but it shouldn't make a difference. Note the lack of spaces. DS is very specific in how entries are made an there are no spaces in its' world.
 
  • Like
Reactions: bjohnrini
Can someone else test this. BI is struggling to run objects:0,dark,combined. My cameras are logging this error.

12/15/2021 7:05:04.922 PM Driveway_Cam Clip: H264 encoder missing (856x480)
 
@105437 Just speculating. Did you stop DS before making changes to your model directory and restart it after you added the new model? What platform are you on CPU or GPU? Is hardware acceleration in use?

@winston_j_allegath Did you do the same, stop DS add the model and restart DS?
 
sebastiantombs said:
@105437 Just speculating. Did you stop DS before making changes to your model directory and restart it after you added the new model? What platform are you on CPU or GPU? Is hardware acceleration in use?

@winston_j_allegath Did you do the same, stop DS add the model and restart DS?
Click to expand...
Yes, I stopped DS while changing the model config. Running the GPU version with a P400 card.
 
@105437 Next thing to try is a reboot. I'm running the V2 combined along with dark, objects:0,combined,dark with not problems at all.

Double check the syntax since DS is very picky that way.

@winston_j_allegath Are you running the CPU or GPU version? It might have gone corrupt and need to be uninstalled and reinstalled. If you do that be sure to delete the entire deepstack directory structure on the C drive before reinstalling.
 
Last edited:
  • Like
Reactions: 105437 and MikeLud1
I am not sure why some are having issues. I have been using general V1.1 on all my cameras.
Just to check if there is an issue with combined V2.0 I just started using combined V2.0 to test if there are any issues with the file sofar V2.0 is working fine for me.
 
  • Like
Reactions: TBurt, 105437 and sebastiantombs
Overnight I tried to test running Combined v2.0 and Dark. Running those two models together still didn't play nice. Today I'll try running Combined v2.0 and the standard Objects model together and see if it's a Dark model issue.

@sebastiantombs You said Combined v2.0 and Dark run fine for you. Would you mind sharing screen shots of your main AI settings and the Trigger > AI settings for your cameras? Thanks!
 
  • Like
Reactions: sebastiantombs
You must log in or register to reply here.
Top Bottom