meituan / YOLOv6

YOLOv6: a single-stage object detection framework dedicated to industrial applications.
GNU General Public License v3.0
5.72k stars 1.03k forks source link

YouLookv6: Shame On You #100

Closed m00nLi closed 2 years ago

m00nLi commented 2 years ago

The so call "YOLOv6": image

https://github.com/ultralytics/yolov5: image

gachiemchiep commented 2 years ago

@m00nLi What made you so mad? From your image, I can see some improvements in inference speed, model size, flops.

image

image

m00nLi commented 2 years ago

Are you sure? 田忌赛马吗?

gachiemchiep commented 2 years ago

@m00nLi Yes I am. I clean up their report numbers a little bit. I only take the inference speed of Speed of (V100, fp32, batch=32). The input size is set into 640

Then the number look like this

image

image

I put the number in Excel so you can comment if you want https://docs.google.com/spreadsheets/d/1AA2d6p7emfuh96Trsle33IPdt0YsYY7QtnXtWsArt6o/edit?usp=sharing

I think authors are still doing their experiments, so bashing them right now is not a good option. I am not Chinese so If you can, please write in English.

ruabliuqiu commented 2 years ago

lose face

You can you up. Don't BiBi if you can't

m00nLi commented 2 years ago

Let me show you: "YOLOv6-n": 11.1 GLOPs, YOLOv5n: 4.5 GFLOPs "YOLOv6-tiny": 36.7 GLOPs, YOLOv5s: 16.5 GFLOPs "YOLOv6-s": 44.2GLOPs, YOLOv5m: 49 GFLOPs Is that clear enough?

m00nLi commented 2 years ago

GFLOPs

Let me show you: "YOLOv6-n": 11.1 GLOPs, YOLOv5n: 4.5 GFLOPs "YOLOv6-tiny": 36.7 GLOPs, YOLOv5s: 16.5 GFLOPs "YOLOv6-s": 44.2GLOPs, YOLOv5m: 49 GFLOPs Is that clear enough?

就你眼尖

I regret your inability

ruabliuqiu commented 2 years ago

GFLOPs

Let me show you: "YOLOv6-n": 11.1 GLOPs, YOLOv5n: 4.5 GFLOPs "YOLOv6-tiny": 36.7 GLOPs, YOLOv5s: 16.5 GFLOPs "YOLOv6-s": 44.2GLOPs, YOLOv5m: 49 GFLOPs Is that clear enough?

就你眼尖

I regret your inability

你若是有能力的话,岂不应该先跑通代码,弄清楚改进点,总结出值得学习的地方,然后再客观评价该算法吗?而不是一上来就喷

gachiemchiep commented 2 years ago

Let me show you: "YOLOv6-n": 11.1 GLOPs, YOLOv5n: 4.5 GFLOPs "YOLOv6-tiny": 36.7 GLOPs, YOLOv5s: 16.5 GFLOPs "YOLOv6-s": 44.2GLOPs, YOLOv5m: 49 GFLOPs Is that clear enough?

That is clear for me. I get your point. They're still in work. Let's them continue your work. If their work still sucks. I will apologize you.