Technography for trajectory prediction. Part of the Surfacing Suspicion project.
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
Ruben van de Ven 332fef2738
Adding ignore file
3 weeks ago
utils Rename to different notebooks, make GPU detector 3 weeks ago
.gitignore Adding ignore file 3 weeks ago
01_detect_objects.ipynb Rename to different notebooks, make GPU detector 3 weeks ago
02_track_objects.ipynb Track objects, and nb to collect and plot tracks 3 weeks ago
03_track_objects_and_collect.ipynb Track objects, and nb to collect and plot tracks 3 weeks ago
04_track_objects_with_deepsort.ipynb Generator pattern & cache of detections pre SORT 3 weeks ago
README.md Initial Notebook with RetinaNet Object Detection 3 weeks ago
poetry.lock Rename to different notebooks, make GPU detector 3 weeks ago
pyproject.toml Rename to different notebooks, make GPU detector 3 weeks ago
sort_abewley.py Add original SORT implementation 3 weeks ago
sort_cfotache.py Rename to different notebooks, make GPU detector 3 weeks ago

README.md

Trajectory Prediction

Use object detector (e.g. RetinaNET, YOLO), and multi-object-tracker (e.g. SORT, DeepSORT) to capture trajectories. Then tinker with Weighted Networks (Hunger 2023) to create predictions of new movements.

Initiate

The project uses poetry. So to set it up:

poetry install

If using python notebook with Jupyter it might be useful to do:

poetry run python -m ipykernel install --user --name trajpred