search

agrimgupta92 / sgan

Code for "Social GAN: Socially Acceptable Trajectories with Generative Adversarial Networks", Gupta et al, CVPR 2018
MIT License
807 stars 259 forks source link

Doubt whether the pooling of SGAN-P really works in collision avoidance #50

Open hunterbobo opened 5 years ago

hunterbobo commented 5 years ago

Hello! First of all. thanks for your great work. But I have a little doubt about whether the pooling really helps avoid collision. I count the number of collision times of SGAN(which doesn't use the pooling) and SGAN-P, specially, the models are downloaded from yours. The threshold distance to judge whether the collision between two person is happened is about 0.08m , which is the minimum distance between two persons in all five datasets. And as show in the following table, SGAN-P doesn't show better collision avoidence performance than SGAN and it seems V-LSTM achieves the best performance. So, really does the pooling help avoid collision? or maybe it is just our garbage imagination.

image

davidglavas commented 4 years ago

Related issue: https://github.com/agrimgupta92/sgan/issues/60

theDebugger811 commented 3 years ago

Hi, This repo might be useful to compare different interaction modules proposed in literature based on collision metrics!