Hasan Enes Şimşek

changed return result.decode('ascii') to return result.decode('utf-8') in interface.py and it works.

https://pjreddie.com/media/files/yolov3.weights https://pjreddie.com/media/files/yolov3-tiny.weights copy these links to web browser ie chrome, downloading will start automatically.

qf has query feature vectors and gf has gallery feature vectors. Multiplication of these two results a new matrix. The element i,j of this matrix represents cosine similarity between ith...

mask = (cls_ids == 20) + (cls_ids == 22) for example 20 for elephants, 22 for zebras full list is here labels = ["person", "bicycle", "car", "motorbike", "aeroplane", "bus", "train",...

What you said is not clear. But as I understand, you are asking the bounding boxes. YOLOv3 detections are good at determining the center point of the bboxes however, due...

here is a good answer https://github.com/ZQPei/deep_sort_pytorch/issues/127#issuecomment-674435481

> Hello, > The evaluation protocol is based on "query" and "gallery", taken from MARS dataset. > But after downloading the dataset, I have only "train" and "test" folders. I...

I trained the model on Market1501 and got really close results and similar epoch-loss results with Pei's ones. But I dont know about Mars.

Thank you for your reply. I think it would be helpful to have your answer in the readme file.