CochDNN

Model code for loading CochDNN auditory models (auditory models with cochleagram front end).

Contains the in-house models used in the paper: Greta Tuckute*, Jenelle Feather*, Dana Boebinger, Josh H. McDermott (2023): Many but not all deep neural network audio models capture brain responses and exhibit correspondence between model stages and brain regions.

Installation and downloading checkpoints.

Required dependencies are specified in setup.py. Install with pip install -e .

Model checkpoints (~14GB) can be downloaded and extracted into the appropriate location with the included script: python download_large_files.py

To test if models load after installation and checkpoint downloading, you can run the test script python tests/test_cochdnn.py. Note: this script will attempt to load the model checkpoints into each architecture for the models without random permutations.

Snippet for loading a model using a build script:

import os
import importlib

# Choose the model that will be loaded
model_dir = 'resnet50_audioset'

build_network_spec = importlib.util.spec_from_file_location("build_network",
                        os.path.join(model_dir, 'build_network.py'))
build_network = importlib.util.module_from_spec(build_network_spec)
build_network_spec.loader.exec_module(build_network)

model, ds = build_network.main()