Derek Gloudemans

icon indicating a website link https://www.linkedin.com/in/derek-gloudemans/

icon company Vanderbilt University Institute for Software Integrated Systems icon location Nashville, TN icon location Computer Science doctoral student at Vanderbilt University. My current studies are in machine learning and computer vision with applications in transportation.

Results 5 comments of Derek Gloudemans

MOT20\train\MOT20-01\gt\gt.txt data format not consistent

To summarize the paper, the columns are as follows: `, , , , , , , , ` where visibility is 1 if the object is completely visible and 0...

MOT20\train\MOT20-01\gt\gt.txt data format not consistent

Not 100% sure, try the paper. I'm guessing they combine truncation (how much of the object is within the frame) and intersection with other objects, possibly taking into account which...

Update README.md

This actually may expose a discrepancy. Ground truth files are specified in 9-value form but I don't know whether the evaluation server still expects a 10-value format detection file. Please...

Should MOTP be higher or lower?

There is a bit of ambiguity in what MOTP means as a metric. (see [here](https://stackoverflow.com/questions/62049456/what-is-multiple-object-tracking-precision-motp)). By one definition (I think the older defininition), MOTP is the average distance between matched...

boxes_data is based on input images coordinate not feature map？

Theoretically, I believe the layers can be used in either order. However, if you are using ROI for FasterRCNN or a similar application, it will be much more computationally efficient...