search

AlexeyAB / darknet

YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
http://pjreddie.com/darknet/
Other
21.65k stars 7.96k forks source link

why yolov3-spp use maxpool size=5,9,13? #2718

Open libralx opened 5 years ago

libralx commented 5 years ago

can i change the size of the maxpool filters?

AlexeyAB commented 5 years ago

Yes, you can. But you should train your own weights-file.

libralx commented 5 years ago

thanks Is there a version of Yolov3-tiny-spp? can i modify the cfg file(yolov3-tiny.cfg）add spp module to use yolov3-tiny-spp by default yolov3-tiny.conv.15 weights @AlexeyAB

libralx commented 5 years ago

Can you tell me how to train the weights-file

发送自 Windows 10 版邮件https://go.microsoft.com/fwlink/?LinkId=550986应用

发件人: Alexey notifications@github.com 发送时间: Tuesday, March 26, 2019 6:56:05 AM 收件人: AlexeyAB/darknet 抄送: blazing; Author 主题: Re: [AlexeyAB/darknet] why yolov3-spp use maxpool size=5,9,13? (#2718)

Yes, you can. But you should train your own weights-file.

― You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/AlexeyAB/darknet/issues/2718#issuecomment-476408724, or mute the threadhttps://github.com/notifications/unsubscribe-auth/AXUFn6wNM30yEeqTf16_08c94uGH0NS2ks5vaVQFgaJpZM4cFi2s.

AlexeyAB commented 5 years ago

Just train as usually: https://github.com/AlexeyAB/darknet#how-to-train-to-detect-your-custom-objects