search

agrimgupta92 / sgan

Code for "Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networks", Gupta et al, CVPR 2018
MIT License
813 stars 261 forks source link

GAN Loss implementation #68

Open zhangpur opened 4 years ago

zhangpur commented 4 years ago

Hi, Thanks for your share of code.

While calculating GAN loss, why do you consider to add a disturbance like uniform(0,7,1.2) or uniform(0,0.3) to the true/fake ground truth, instead of purely 1 or 0?

Thanks for your help in advance.