FasterNet icon indicating copy to clipboard operation
FasterNet copied to clipboard
verified publisher icon JierunChen

some question

Open jackiezhang11 opened this issue 2 years ago • 4 comments

Can I use 416416 picture as input？ what about 448448 in detection without pre-training model?Could fasternet as YOLOV3 backbone?

jackiezhang11 avatar Apr 27 '23 06:04 jackiezhang11

@jackiezhang11 Hi, both 416 and 448 can be the input size. The detection performance would degrade without pre-training. And yes, FasterNet should be applicable as a backbone for YOLOV3.

JierunChen avatar Apr 29 '23 00:04 JierunChen

Thank you for quickly apply.Could you give me an idea to make FasterNet as YOLOV3's backbone.Is it altering network in Pytorch and altering cfg file in darknet?I don't konw is it right.If you konw,Could you tell me details to alter?

jackiezhang11 avatar Apr 29 '23 02:04 jackiezhang11

Use Fasternet as backbone would improve accuracy.Could you tell me the reason ?such as compared with YOLOV3.

jackiezhang11 avatar May 13 '23 03:05 jackiezhang11

Use Fasternet as backbone would improve accuracy.Could you tell me the reason ?such as compared with YOLOV3.

Hi, I have been following how to replace FasterNet with ResNet recently. Have you implemented it yet? How is the effect? Can we discuss it? Thank you very much.

zhaigz avatar Nov 12 '24 02:11 zhaigz