search

dougsm / ggcnn

Generative Grasping CNN from "Closing the Loop for Robotic Grasping: A Real-time, Generative Grasp Synthesis Approach" (RSS 2018)
BSD 3-Clause "New" or "Revised" License
499 stars 140 forks source link

grasp quality #6

Closed wjydou closed 5 years ago

wjydou commented 5 years ago

Hello,I have completed the training, but how do I test the grasp quality ?

dougsm commented 5 years ago

Hello, after you have finished the training there is a script evaluate.py that will test the network. There are some parameters you can set in the file before running it, e.g. the path to your trained models and how many grasps to evaluate per image.

wjydou commented 5 years ago

@dougsm Yes, I have already run evaluate.py, and the result of the output is a txt text.Some of the numbers of such output, I don’t know what that mean.So, can I detect the images that I have collected without running the robot to grab?