Prerequisite

  • Python 3
  • OpenCV (i.e. able to import cv2)

Data:

  • 4.5G 18-01-01 -- Eric
  • 7.3G 18-01-02 -- Laetitia
  • 3.2G 18-02-01 -- Zaid
  • 4.8G 18-02-02 -- Kye
  • 5.8G 18-02-03 -- Ruchir
  • 2.6G 18-03-01 -- Avishi
  • 4.1G 18-03-02
  • 2.3G 18-03-03
  • 773M 18-03-04

Steps

  1. Please scp the corresponding folder (under /data/icu/raw) and scripts (~/untar.py and ~/data_annotate_NIPS18) to your local machine (edited)

  2. Prepare folder structure on your local machine:

  • create a folder named e.g. labeling_party
  • cd into labeling_party
  • create a subfolder named raw
  • put your data (e.g. 18-03-01) under labeling_party/raw
  • put scripts (untar.py and data_annotate_NIPS18) under labeling_party
  1. Find the keymap for your machine: please run data_annotate_NIPS18/annotate/test_key_map.py, press keys following the prompts, and update the values in dictionary KEYMAP in data_annotate_NIPS18/annotate/utils.py.

  2. To start the labeling interface:

  • cd data_annotate_NIPS18
  • Modify scripts/annotate.sh: update date and sensor.
  • run ./scripts/annotate.sh, or directly run python -m annotate.run --date={your_date} --sensor={your_sensor}, e.g. python -m annotate.run --date=18-01-01 --sensor=171