The paper is really interesting, good job! I tried using the social gan model and found that the predictions it produces often contain collisions (pedestrians collide). Experiments show that the models without the pooling module tend to have just as much or less collisions than the models with the pooling module. This contradicts the rationale behind the pooling module which is supposed to avoid collisions.
The experiment:
Each situation consists of a sequence of frames, for each frame I compute the euclidian distance between all pedestrians, if any of these distances is below a threshold then I classify the situation that this frame belongs to as containing a collision. Each of the following four charts describes five of the in total twenty models that you provided via the download script:
Interestingly, the pooling module doesn't seem to reduce the number of collisions (compare left two and right charts).
See a bit more details about the experiments here.
The paper is really interesting, good job! I tried using the social gan model and found that the predictions it produces often contain collisions (pedestrians collide). Experiments show that the models without the pooling module tend to have just as much or less collisions than the models with the pooling module. This contradicts the rationale behind the pooling module which is supposed to avoid collisions.
The experiment:
Each situation consists of a sequence of frames, for each frame I compute the euclidian distance between all pedestrians, if any of these distances is below a threshold then I classify the situation that this frame belongs to as containing a collision. Each of the following four charts describes five of the in total twenty models that you provided via the download script:
Interestingly, the pooling module doesn't seem to reduce the number of collisions (compare left two and right charts).
See a bit more details about the experiments here.