knandanan

Results 4 comments of knandanan

@Sakuralala I also need clarification regarding the same . Did you find a method to go about prediciting without scale and center?

@pmalatyn to clarify, the pipeline has two back-to-back gvadetect where the first gvadetect runs on full frame and the second gvadetect runs on inference-region=roi-list and object-class=person. For the detections from...

> understood, but gvadetect sets parent_id only if you use roi-list (and not full-frame). SO in your case the 1st pipeline is running on full frame so there is nothing...