xincoder

@LYWZ12138 thank you for asking, their code is still accessible at https://github.com/rayguan97/Spectral-Trajectory-Prediction. Thanks.

@fgqile Thank you for your interest in our work. As we mentioned in our GRIP paper (section V.A Datasets), we use the data preprocessing code released by Nachiket Deo et...

@KevinDu1 首先非常感谢你对我们工作的支持。这个表现具体是什么原因比较难判断，不过有可能是数值稳定性的问题，建议检查以下几个地方: (1). 确保输入到模型的数据是对的。我分享出来的code是在Apollo数据集上的，dataloader范围的dimension比较多，在main.py [line94] 选择需要输入到模型的dimension，main.py[line99]计算velocity。可以double check一下输入是否正确。(2). 虽然固定了输入车辆的个数N， 但是在处理数据的时候，我把所有valid的objects排在了前边，参见data_process.py[line47] process_data()函数，[line49]取到有效的object个数num_visible_object。这一点在处理数据的时候也可以注意一下。(3).有可能Argoverse数据及上的数据范围比较特殊，可以尝试做一下rescale。我的code里实际上实现了这一功能，main.py [27-28] 对应修改max_x 和 max_y。 (4). 可以尝试在训练时候添加gradient clipping，以避免梯度异常。

Hi @lylyjy , thank you for your interest in our work. Please check your data_process.py, you should correct [line 8] so that "prediction_train/*.txt" and "prediction_test/*.txt" are located in your [line...

> > Hi @lylyjy , thank you for your interest in our work. Please check your data_process.py, you should correct [line 8] so that "prediction_train/_.txt" and "prediction_test/_.txt" are located in...

@lylyjy no, you do not need to do such a modification. "*" means any character (or string) here. Please refer to the glob official document for more information. Thanks.

Hi @poojitharamachandra , thank you for your interest. The GRIP works in 3D space. If you look at the input dimensions, it does not have a specific design to only...

1. The "mat_hop" is passed into Class Graph() in layers/graph.py. It is used in GRIP/layers/graph.py [line 21], np.linalg.matrix_power(A, d). In Graph Theory, adjacency matrix powers (A^hop_step) gives the number of...

@poojitharamachandra , we only take objects that appear at the last observed time step into account (data_process.py line48) because objects that do not appear at this time step have a...

> @xincoder thank you very much > How do You ensure that the predicted vehicle trajectories do not overlap? (I have this problem in my usecase) @poojitharamachandra It is a...