may i ask if i train the model,should i use RGB-D images or just RGB images?

[Jonho111]

hello,i have read the paper(Real-Time Grasp Detection Using Convolutional Neural Networks),and the paper saied that uses the RGB-D image to train the model,and may i ask the code in this project,should i use RGB-D images or RGB images ,or both of them is ok?thank you very much, this problem is very important to me!!