Junru Gu
Junru Gu
It seems that `model_recover_path` does not appear in the README
What is your evaluation command? During evaluation, you should set `model_recover_path` to the correct ${OUTPUT_DIR}/model_save/model.16.bin.
Sorry that I'm not able to answer this question. It seems like the issue of the multiprocessing module. Have you try some simple example codes to run multiprocessing?
It is updated after calculating scores of dense goal https://github.com/Tsinghua-MARS-Lab/DenseTNT/blob/a0e3b8a51aecf9f9046db4fb72e2793684c96e69/src/modeling/decoder.py#L158
This is result of single trajectory prediction. Performing evaluation will get results of multi-trajectory prediction.
Optimizing strategy during evaluation can be found in the paper. `MRminFDE` is used to control the ratio of optimizing miss rate. For example, "0.5" means optimizing miss rate and minFDE...
We use argoverse library to calculate metrics. See https://eval.ai/web/challenges/challenge-page/454/evaluation
You may refer to https://github.com/Tsinghua-MARS-Lab/M2I
We found that different training strategies achieve similar performance. We recommend enhancing the encoder for agents and lanes to improve the performance.
1. Loss for lane can be found in 'lane_scoring' module (in our updated code) 2. New detailed description for this function may be helpful 3. Scores of heatmap indicate probability...