Courtois-Neuromod Imagebank (CIB)

Code to inventorize and see stimuli distribution across categories

Use the 'read_json' function in 'cib_utils.py' to import 'cib_inv.json' in a Python IDE as a dict.

A graphic variable explorer such as provided in Spyder is helpful -
- Currently unavailable: Index-value pairs are not yet recognized by pandas
- see TO-DO*

[_] Make cib_inv.json more readable
[_] Use COCO and SUN databases to complete the 'places' top category
[_] Validate category distribution asymetries according to either or both:

A) Neuronal sources

B) Human validated image databases

C) Artificially (DNN) validated image datasets
- [_] Re-organize categories with less than 20 images into more basic classification levels (eg. bird talons all in the same category, regardless of bird species)
[_] Re-implement WN compatibility In Spyder 4.x, wn.synset objects are returned as 'metacontainer' objects, which are too large to load in a dataframe as in previous versions.
- [_] Synset disambiguation for items in 'place' category

Name		Name	Last commit message	Last commit date
Latest commit History 14 Commits
__pycache__		__pycache__
old_docs		old_docs
scripts		scripts
CogNeuroModels.xlsx		CogNeuroModels.xlsx
License.txt		License.txt
README.md		README.md
all_synsets.xlsx		all_synsets.xlsx
cib_inv.json		cib_inv.json
cib_inv_indent36.json		cib_inv_indent36.json
cib_map_short.json		cib_map_short.json
cib_utils.py		cib_utils.py
image_tags.csv		image_tags.csv
load_cib.py		load_cib.py
load_cib_short.py		load_cib_short.py
not_20_imgs.xlsx		not_20_imgs.xlsx
object_concepts_plus.xlsx		object_concepts_plus.xlsx
sources.csv		sources.csv