search

I24-MOTION / VT_tools

Toolbox for creating spatiotemporal speed fields and generating virtual trajectories within these fields. Virtual trajectories are useful for analytics because they are smooth, complete, and can be generated at specific and uniform times. This code accompanies an article presented at IEEE-FISTS 2024 (Ji 2024).
https://i24motion.org
Other
2 stars 2 forks source link

Speeding up the data processing file (about 300 times faster) and hyper parameter adjustments #2

Open seungwooham opened 1 month ago

seungwooham commented 1 month ago

First of all, thanks for providing us the baseline code.

This pull request contains an enhancement in data processing ipynb file. The code before took long time to search for a specific value and process it row-by-rows. Here, the adjustment was made based on the assumption the pre-processed file has 2d array with x-position and time stamp, starting from 0. (which is intuitive.)

Based on this, the time consumption for data processing has now been reduced by 1/300.

It also has some adjustments to the hyperparameters based on the comparison with the original paper.

barbourww commented 1 month ago

@seungwooham Thank you for the submission! We will start reviewing this immediately. The end result might be that we support two input versions if needed in order to allow backwards compatibility for a while.

junyi9 commented 2 weeks ago

@seungwooham Thanks for the submission, we'll incorperate your code logic in our next release.