For each image collection (e.g., corresponding to the stills from one video), there must be the following files:
images/
- directory with images that were labeled
conf.js
- Javascript objects representing labeling environment when these images were labeled. Defines the available type labels and subtype labels. Also names and describes the modifier.
img_arr_init.js
- Javascript array of image filenames and labels before any labeling happened (all images unlabeled; all images unseeen). This file can be created with the getstills
Python module. It is for initializing the Keystroke Labeler utility with the images in the images/
directory.
If labeling has taken place, and data has been exported, the following files may be associated with the image collection:
img_arr_prog.js
- Javascript array of images and labels after labeling. This file can be used to re-start the Keystroke Labeler utility with labels already entered.
img_labels.csv
- CSV export of image labels. Contains exactly the same data as in img_arr_prog.js
.
The tabular data has seven columns, as labeled in the CSV file.
-
filename (string) - the filename of the image to be labeled
-
seen (bool) - indicates whether the item has been seen
-
type label (char) - indicates the type label, if any, of seen items
-
subtype label (char) - indicates the subtype label, if any, of items with type labels
-
modifier (bool) - indicates whether the label has the "modifier" status
-
transcript (string) - displayed and saved if included in
img_arr_init.js
orimg_arr_prog.js
; not editable within labeler -
note (string) - displayed and saved if included in
img_arr_init.js
orimg_arr_prog.js
; not editable within labeler