FairMOT icon indicating copy to clipboard operation
FairMOT copied to clipboard

FaceMOT

Open zengwb-lx opened this issue 5 years ago • 5 comments

I trained a face tracking model with HRnet18, and the tracking performance was good (it could be tracked back within 1S after being lost)But not fast enough in the 2070S(about 30fps)。could you give me some Suggestions for switching to a lighter network

zengwb-lx avatar Sep 14 '20 09:09 zengwb-lx

You can use ResNet-34, which is much faster. The sh file of training ResNet-34 can be found in the experiments folder.

ifzhang avatar Sep 14 '20 11:09 ifzhang

@zengwbz Hey can you please suggest you're training process and scripts if possible. Also, which face tracking dataset did you use?

Ashwin-Ramesh2607 avatar Sep 25 '20 09:09 Ashwin-Ramesh2607

Face detection data and face recognition data are combined together Wider-Face and Casia-webface, using mobilenet_fpn backbone and Multi-scale inputs provide better performance

zengwb-lx avatar Sep 25 '20 09:09 zengwb-lx

@zengwbz any chance you can release the trained face tracking model?

kevinwei30 avatar Oct 22 '20 01:10 kevinwei30

@Ashwin-Ramesh2607 @zengwb-lx Hi is there any face tracking model ?

MyraBaba avatar Oct 13 '21 09:10 MyraBaba