Skip to content
/ mAP Public
forked from Cartucho/mAP

mean Average Precision - This code evaluates the performance of your neural net for object recognition.

License

Notifications You must be signed in to change notification settings

HeejoonLee/mAP

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

mAP for the 2nd SK AI Competition

ref: https://github.com/Cartucho/map

Modifications

  • Changed IoU from 0.5 to 0.6
  • Use specified class weights to calculate mAP:
        SK_AI_class_weights = {
            "normal": 0.17,
            "unscrewed_red": 0.23,
            "rusty_yellow": 0.17,
            "rusty_red": 0.19,
            "unscrewed_yellow": 0.19,
            "none": 0.05
        }
        sum_AP += ap * SK_AI_class_weights[class_name]

Usage

python3 main.py

Notes

Make sure the input directory hierarchy and input file format match that of the original source.

TODO

Create a script to automatically create input directory structure with correct file formats from model output files

About

mean Average Precision - This code evaluates the performance of your neural net for object recognition.

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • Python 100.0%