Preprocessing Haggling DB from the Panoptic Studio Dataset
- gathering_panoptic.py
This file loads all haggling json files from /yourPath/panoptic-toolbox/data_haggling, which was downloaded from domedb, and saves them as pkl files.
Similarly, run the following for face, and hands:
gathering_panopticDB_face.py
gathering_panopticDB_hand.py
- process_hagglingDB.py This code does some additional processing for each haggling sequence, as follows:
-
- Align the startFarmes for the people in the same group
-
- Separate the name of pkl file, to have a single group only
Similarly, run the following for face, and hands:
process_hagglingDB_face.py
process_hagglingDB_hand.py