Hi,
Can the community guide me if I am missing anything?
I use an E340 Floodlight camera and a Homebase 3 for a year now.
The system detects dozens of false positives a day when the sun is behind a tree and shadows fall into the defined detection area. It also detects chairs as pets.
I can understand false positives for a newly deployed system, however I expect the system, which markets itself as an AI learning system, to learn from labeling over 3500 false positive events in the past year.
Has someone managed to train the system to actually learn from these manual labeling?
What am I missing? Why doesn’t the system learn? I spent hours on labeling over 3500 false positive events and there is zero progress.
Firmware was always updated to the latest version as soon as it was released.
Thanks in advance for any guidance.
T.