Open agn-7 opened 5 years ago
I realized that it's relevant to the classifier because of their clusters (stand persons) are true.
Hi @agn-7 , As you pointed out, maybe this is due to the difference of the human shape in real and simulated data. You may track such still persons by setting "enable_classification" to false. It makes the human classifier disabled.
Hi @koide3
No, this not due to the human shape, because when I move up simulated humans this repo detect them, with enable_classification
true or false, but when simulated humans are fixed with enable_classification
true it cannot detect them, and with enable_classification
false, it detects them. But with enable_classification
false (clustering only) the False-Positive value increased up.
@agn-7, Hi Were you able to solve the issue of static pedestrian? I am also facing the same issue. In my pointclouds, when people are static its failing to detect. Once they started moving its started to detect. Nithin
I did an experimental person detection on a simulator which its persons are stand/fixed, so I realized that this repo does not find the fixed person and when I move persons it detected them.
Is there any parameter to detect stood persons or if not, can I change a section of code to detect them?