search

WongKinYiu / yolor

implementation of paper - You Only Learn One Representation: Unified Network for Multiple Tasks (https://arxiv.org/abs/2105.04206)
GNU General Public License v3.0
1.98k stars 524 forks source link

How to predict all the objects in an image that contains handreds of objects? #150

Closed hughwyc closed 2 years ago

hughwyc commented 2 years ago

I use the YOLOR P6 to predict an image like this:

003

The command that I run is as follows: python detect.py --source inference/images/003.jpg --cfg cfg/yolor_p6.cfg --weights yolor_p6.pt --conf 0.25 --img-size 1280 --device 0

But only few of objets in the image was detected ( 23 persons, 5 handbags ) , the input image size is 1920x1280. Does anyone konws how to get an ideal result by changing some parameters in the code?

kadmor commented 2 years ago

You can try to reduce --conf

hughwyc commented 2 years ago

thanks, it works.

You can try to reduce --conf