Closed ozgurozc closed 3 years ago
Yup, cool use-case. I'd love to see some pictures if possible
Use the new union
value in same_model_strategy
If you are not part of the ZM slack channel, would be great if you could join - love to hear more of what you are doing/use-case
app:6.1.16, pyzm:0.3.38, ES:6.1.16 is old. You will need to update to latest versions to use union.
Thank you for your reply , I had been mastering your notification server quite a while in the pandemic lockdown times , loved the way it is coded, It is great job.
Yup, cool use-case. I'd love to see some pictures if possible Use the new
union
value insame_model_strategy
See
If you are not part of the ZM slack channel, would be great if you could join - love to hear more of what you are doing/use-case
I had joined the slack channel and will love to share details.
Before you create an issue, please make sure you have read the README. If you are asking about the object detection part, I don't provide support for it unless you've tried hard enough.
Event Server version app:6.1.16, pyzm:0.3.38, ES:6.1.16 , OpenCV:4.5.1-dev
Are you using MLAPI? (Y/N) NO
Question:
I have a set of cameras detecting cars, tracks and persons and alarm to me with zmninja successfully, I had added a yolo4 pre-trained mask detection model and get it worked in zmeventnotification server. Is it possible to add two models under object section of objectconfig.ini? since it is either detecting person or detecting mask ,when I am running the model it is always bypassing the results of mask detection eventough it is successfully detecting. Is it possible to write the results of two models in the same objdetect.jpg and object.json file like it is doing with object, face and Alpr. I had tried the "union" as same_model_strategy but did not worked.
Below I had managed to run the two models in the same detection sequence . objectconfigini.txt
Below you can see the result debug_match_detection_results_with_person_witout_mask_detection_results.txt
Below you can see the the successful writing of mask detection results when I disabled the yolo4 object detection. debug_mask_detection_results_to_json.txt