I have a couple of high-resolution cameras that are set up to detect people only, but I'm getting multiple hits for our dogs in the backyard, and one cat that came into view in our front yard.
You can clearly see four legs, so I'm not sure if the algorithm has just never learned that four legs = not human, or if there is some other issue to squelch these false hits.
Anyone have any experience here? Can I submit the video to sighthound so they can use machine learning and determine that these aren't people?
