Inhwan Bae

Hi @miao02830! Would you like to try [old version of ENet-SAD codebase](https://github.com/InhwanBae/ENet-SAD_Pytorch/tree/f4e07c6298cafffbfd33fbd006ebffeec99e7432)? The results reported in this repository are trained and evaluated using the previous version of the code. Note...

Hi @Pradur241, Of course! You can extract the group indices without a predictor using [`line 212`](https://github.com/InhwanBae/GPGraph/blob/49d86eca04aa72095f949609f1a0fc731a6476e0/model_groupwrapper.py#L211-L212) in the [`model_groupwrapper.py`](https://github.com/InhwanBae/GPGraph/blob/49d86eca04aa72095f949609f1a0fc731a6476e0/model_groupwrapper.py#L211-L212). I have checked that any predictors other than trajectories or directly...

@Pradur241 The value of `self.th` is a learnable parameter that divides whether two people are in the same group or not, based on the feature distance between them. In easy...

Hi, @Luo624! I used a convex hull algorithm to group the members into regions for visualization. I hope the code snippet below is helpful to you. ```python from scipy.spatial import...

Hi @Dong3759! In addition to the video for trajectory visualization, it's necessary to redesign the dataloader to incorporate the homography matrix, frames, and more. The visualization source code contains several...

Hi @Dong3759, I'm sharing the visualization source code with you via a [**`public repository`**](https://github.com/InhwanBae/ETH-UCY-Trajectory-Visualizer) because I haven't received any mail from you. Check it out! If you need any other...

Hi @12num, Exactly. Before starting the training, all trajectories within the training dataset are decomposed using SVD to construct ET space. It will be much easier to understand if you...

Hi @Jasmine302, Your error seems to be the same as the one I noticed when I recently reran the source code. As mentioned in [issue#1](https://github.com/InhwanBae/GraphTERN/issues/1), it might help to use...

Thanks for your interest in my work! In my experience, when training on the ETH set, early stopping between 16 and 32 epochs significantly improves performance. Even so, FDE 1.4...

Hi @Pranav-chib, I'll share the results of my tests from the past week and some interesting findings with you. First, I was able to reproduce the table in the paper...