search

AlexeyAB / darknet

YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
http://pjreddie.com/darknet/
Other
21.68k stars 7.96k forks source link

Compare inference time between YOLO-EfficientNetB0 with YOLO-tiny #4457

Open loi-nguyen-khanh opened 4 years ago

loi-nguyen-khanh commented 4 years ago

Hi all,

I try to use object detection with YOLO-EfficientNetB0 and YOLO-tiny model. As I know, YOLO-tiny has 5.6Bn FLOPS, YOLO-EfficientNetB0 has only 3.7Bn FLOPS. But when test a video. with YOLO-tiny , i can run with ~220FPS YOLO-ENET, i run with only ~60FPS

pawan9 commented 4 years ago

please see https://github.com/AlexeyAB/darknet/issues/4447#issuecomment-561611786

Jack-ouc commented 4 years ago

Hello,I searched a lot about the speed of EfficientNet, because I implement it on GTX1060 and feel it slower than darknet(in yolov3),some one said it is really slow,but if you can run it with 60FPS , I think it is nice, so I want to ask you , did you run YOLO-ENET with 60FPS really? And what's the environments?

loi-nguyen-khanh commented 4 years ago

Hello,I searched a lot about the speed of EfficientNet, because I implement it on GTX1060 and feel it slower than darknet(in yolov3),some one said it is really slow,but if you can run it with 60FPS , I think it is nice, so I want to ask you , did you run YOLO-ENET with 60FPS really? And what's the environments?

Hi, I run on computer that has config as bellow: CPU: i7-9700K GPU: GeForce RTX 2080Ti

Jack-ouc commented 4 years ago

Hello,I searched a lot about the speed of EfficientNet, because I implement it on GTX1060 and feel it slower than darknet(in yolov3),some one said it is really slow,but if you can run it with 60FPS , I think it is nice, so I want to ask you , did you run YOLO-ENET with 60FPS really? And what's the environments?

Hi, I run on computer that has config as bellow: CPU: i7-9700K GPU: GeForce RTX 2080Ti

Thank you!