yzqin / s4g-release

S4G: Amodal Single-view Single-Shot SE(3) Grasp Detection in Cluttered Scene
35 stars 5 forks source link

Changing model parameters #7

Closed kimanton closed 1 year ago

kimanton commented 1 year ago

A question regarding the model. In file curvature model, if I want to change model parameters, does that mean I can't use your pretrained weights?

yzqin commented 1 year ago

Hi @kimanton,

Yes, if you change the neural network model parameters under model, then the the architecture is changed and you can not use the pre-trained weights. It is the same case for all learning based method and neural networks.

Sorry that you can not change these parameters if you want to use the pre-trained models. But you can change all the post-processing parameters code that is not related to neural network.

kimanton commented 1 year ago

Thank you for explanation