Object Detection Part 6: The Hungarian Matching Algorithm, Tracking, Bounding Box Matching

In this episode, we delve into the fascinating world of optimization algorithms with a focus on the Hungarian Algorithm, also know as the Kuhn–Munkres algorithm or the Munkres assignment algorithm, which is a powerful optimization algorithm that solves the assignment problem in polynomial time. It is widely used in computer vision and object detection tasks to match the objects between the consecutive frames of a video.

*Related Videos*
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

*Contents*
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
00:00 - Intro Hungarian Algorithm
00:38 - Object Tracking with Hungarian Matching
02:01 - Step 1 - Cost Matrix
02:20 - Step 2 - Apply the Hungarian Algorithm
03:54 - Step 3 - Perform the assignment
04:06 - Hungarian Algorithm Drawbacks
04:22 - Outro

*Follow Me*
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬

*Channel Support*
▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬▬
The best way to support the channel is to share the content. ;)

If you'd like to also support the channel financially, donating the price of a coffee is always warmly welcomed! (completely optional and voluntary)
► Bitcoin (BTC): 3C6Pkzyb5CjAUYrJxmpCaaNPVRgRVxxyTq
► Ethereum (ETH): 0x9Ac4eB94386C3e02b96599C05B7a8C71773c9281
► Cardano (ADA): addr1v95rfxlslfzkvd8sr3exkh7st4qmgj4ywf5zcaxgqgdyunsj5juw5
► Tether (USDT): 0xeC261d9b2EE4B6997a6a424067af165BAA4afE1a

#objectdetection #computervision #hungarianmatching #tracking

[AndyHOTlife]

It seems to me like the assignment (0, 1) (1, 0) and (2, 2) would have also been possible, giving an incorrect result. This explanation does not make a lot of sense

[pewpewhuang]

Is this algorithm practical for real time applications? or other method are used more often?

I still see some papers mention it in pair with YOLO.