model-vs-human
Benchmark your model on out-of-distribution datasets with carefully collected human comparison data (NeurIPS 2021 Oral)
File Explorer
Download Latest Version (.zip)- header.png
- airplane.png
- bear.png
- bicycle.png
- bird.png
- boat.png
- bottle.png
- car.png
- cat.png
- chair.png
- clock.png
- colorbar.pdf
- dog.png
- elephant.png
- keyboard.png
- knife.png
- oven.png
- response_icons_horizontal.png
- response_icons_vertical.png
- response_icons_vertical_reverse.png
- truck.png
- all_noise-generalisation_stimuli.png
- all_nonparametric_stimuli.png
- evaluate.py
- plotting_definition.py
- benchmark_figures.tex
- benchmark_table_accuracy.tex
- benchmark_table_humanlike.tex
- benchmark_tables.tex
- consistency_vs_accuracy.tex
- error_consistency_lineplots.tex
- error_consistency_matrices.tex
- noise_generalisation.tex
- nonparametric_accuracy.tex
- shape_bias.tex
- example-report.pdf
- neurips.sty
- report.pdf
- report.tex
- CODE_LICENSE
- LICENSES_OVERVIEW.md
- MODEL_LICENSES
- __init__.py
- base.py
- create_dataset.py
- dataloaders.py
- dataset_converters.py
- decision_mappings.py
- experiments.py
- imagenet.py
- info_mappings.py
- noise_generalisation.py
- registry.py
- sketch.py
- stylized.py
- texture_shape.py
- __init__.py
- evaluate.py
- imagenet_labels.txt
- metrics.py
- __init__.py
- categories.txt
- human_categories.py
- plotting_helper.py
- wordnet_functions.py
- __init__.py
- robust_models.py
- __init__.py
- kerasnet.py
- pytorchnet.py
- __init__.py
- imagenet_classes.py
- imagenet_templates.py
- __init__.py
- pycontrast_resnet50.py
- __init__.py
- texture_shape_models.py
- __init__.py
- cores.py
- __init__.py
- gdrive.py
- mlayer.py
- modules.py
- __init__.py
- simclr.py
- __init__.py
- __init__.py
- model_zoo.py
- __init__.py
- build_model.py
- model_zoo.py
- tf_hub_model_url.py
- __init__.py
- base.py
- pytorch.py
- tensorflow.py
- __init__.py
- registry.py
- __init__.py
- analyses.py
- colors.py
- decision_makers.py
- plot.py
- __init__.py
- cli.py
- constants.py
- model_evaluator.py
- utils.py
- version.py
- colour_subject-01_session_1.csv
- colour_subject-02_session_1.csv
- colour_subject-03_session_1.csv
- colour_subject-04_session_1.csv
- contrast_subject-01_session_1.csv
- contrast_subject-02_session_1.csv
- contrast_subject-03_session_1.csv
- contrast_subject-04_session_1.csv
- cue-conflict_subject-01_session_1.csv
- cue-conflict_subject-02_session_1.csv
- cue-conflict_subject-03_session_1.csv
- cue-conflict_subject-04_session_1.csv
- cue-conflict_subject-05_session_1.csv
- cue-conflict_subject-06_session_1.csv
- cue-conflict_subject-07_session_1.csv
- cue-conflict_subject-08_session_1.csv
- cue-conflict_subject-09_session_1.csv
- cue-conflict_subject-10_session_1.csv
- edge_subject-01_session_1.csv
- edge_subject-02_session_1.csv
- edge_subject-03_session_1.csv
- edge_subject-04_session_1.csv
- edge_subject-05_session_1.csv
- edge_subject-06_session_1.csv
- edge_subject-07_session_1.csv
- edge_subject-08_session_1.csv
- edge_subject-09_session_1.csv
- edge_subject-10_session_1.csv
- eidolonI_subject-01_session_1.csv
- eidolonI_subject-02_session_1.csv
- eidolonI_subject-03_session_1.csv
- eidolonI_subject-04_session_1.csv
- eidolonII_subject-01_session_1.csv
- eidolonII_subject-02_session_1.csv
- eidolonII_subject-03_session_1.csv
- eidolonII_subject-04_session_1.csv
- eidolonIII_subject-01_session_1.csv
- eidolonIII_subject-02_session_1.csv
- eidolonIII_subject-03_session_1.csv
- eidolonIII_subject-04_session_1.csv
- false-colour_subject-01_session_1.csv
- false-colour_subject-02_session_1.csv
- false-colour_subject-03_session_1.csv
- false-colour_subject-04_session_1.csv
- high-pass_subject-01_session_1.csv
- high-pass_subject-02_session_1.csv
- high-pass_subject-03_session_1.csv
- high-pass_subject-04_session_1.csv
- low-pass_subject-01_session_1.csv
- low-pass_subject-02_session_1.csv
- low-pass_subject-03_session_1.csv
- low-pass_subject-04_session_1.csv
- phase-scrambling_subject-01_session_1.csv
- phase-scrambling_subject-02_session_1.csv
- phase-scrambling_subject-03_session_1.csv
- phase-scrambling_subject-04_session_1.csv
- power-equalisation_subject-01_session_1.csv
- power-equalisation_subject-02_session_1.csv
- power-equalisation_subject-03_session_1.csv
- power-equalisation_subject-04_session_1.csv
- rotation_subject-01_session_1.csv
- rotation_subject-02_session_1.csv
- rotation_subject-03_session_1.csv
- rotation_subject-04_session_1.csv
- silhouette_subject-01_session_1.csv
- silhouette_subject-02_session_1.csv
- silhouette_subject-03_session_1.csv
- silhouette_subject-04_session_1.csv
- silhouette_subject-05_session_1.csv
- silhouette_subject-06_session_1.csv
- silhouette_subject-07_session_1.csv
- silhouette_subject-08_session_1.csv
- silhouette_subject-09_session_1.csv
- silhouette_subject-10_session_1.csv
- sketch_subject-01_session_1.csv
- sketch_subject-02_session_1.csv
- sketch_subject-03_session_1.csv
- sketch_subject-04_session_1.csv
- sketch_subject-05_session_1.csv
- sketch_subject-06_session_1.csv
- sketch_subject-07_session_1.csv
- stylized_subject-01_session_1.csv
- stylized_subject-02_session_1.csv
- stylized_subject-03_session_1.csv
- stylized_subject-04_session_1.csv
- stylized_subject-05_session_1.csv
- uniform-noise_subject-01_session_1.csv
- uniform-noise_subject-02_session_1.csv
- uniform-noise_subject-03_session_1.csv
- uniform-noise_subject-04_session_1.csv
- .gitignore
- README.md
- setup.cfg
- setup.py
# Installation Guide
1. Get the code
git clone https://github.com/bethgelab/model-vs-human
Downloads the entire project code from GitHub to your computer.
cd model-vs-human
Moves into the project folder you just downloaded.
2. Python
Easy RecommendedPrerequisites
pip install -e .
Installs the Python libraries listed in requirements.txt (or similar).
If it runs without errors and prints output in the terminal, it worked.
Pulled directly from this repo's README.
// repository documentation
Was this content helpful?
(0 ratings)