Newer
Older
.. Copyright © 2022 Idiap Research Institute <contact@idiap.ch>
..
.. SPDX-License-Identifier: GPL-3.0-or-later
=====================
Preset Configurations
=====================
.. _mednet.libs.classification.config:
------------------------------------
Classification Preset Configurations
------------------------------------
This module contains preset configurations for baseline CNN architectures and
DataModules in a classification task.
.. _mednet.libs.classification.config.models:
André Anjos
committed
Pre-configured Models
André Anjos
committed
Pre-configured models you can readily use.
André Anjos
committed
:toctree: api/config.models
mednet.libs.classification.config.models.alexnet
mednet.libs.classification.config.models.alexnet_pretrained
mednet.libs.classification.config.models.densenet
mednet.libs.classification.config.models.densenet_pretrained
mednet.libs.classification.config.models.densenet_rs
mednet.libs.classification.config.models.logistic_regression
mednet.libs.classification.config.models.mlp
mednet.libs.classification.config.models.pasa
.. _mednet.libs.classification.config.datamodules:
André Anjos
committed
DataModule support
André Anjos
committed
Base DataModules and raw data loaders for the various databases currently
supported in this package, for your reference. Each pre-configured DataModule
André Anjos
committed
can receive the name of one or more splits as argument to build a fully
functional DataModule that can be used in training, prediction or testing.
André Anjos
committed
:toctree: api/config.datamodules
mednet.libs.classification.config.data.hivtb.datamodule
mednet.libs.classification.config.data.indian.datamodule
mednet.libs.classification.config.data.montgomery.datamodule
mednet.libs.classification.config.data.montgomery_shenzhen.datamodule
mednet.libs.classification.config.data.montgomery_shenzhen_indian.datamodule
mednet.libs.classification.config.data.montgomery_shenzhen_indian_padchest.datamodule
mednet.libs.classification.config.data.montgomery_shenzhen_indian_tbx11k.datamodule
mednet.libs.classification.config.data.nih_cxr14.datamodule
mednet.libs.classification.config.data.nih_cxr14_padchest.datamodule
mednet.libs.classification.config.data.padchest.datamodule
mednet.libs.classification.config.data.shenzhen.datamodule
mednet.libs.classification.config.data.tbpoc.datamodule
mednet.libs.classification.config.data.tbx11k.datamodule
André Anjos
committed
.. _mednet.libs.classification.config.datamodule-instances:
André Anjos
committed
André Anjos
committed
Pre-configured DataModules
^^^^^^^^^^^^^^^^^^^^^^^^^^
André Anjos
committed
André Anjos
committed
DataModules provide access to preset pytorch dataloaders for training,
validating, testing and running prediction tasks. Each of the pre-configured
DataModule is based on one (or more) of the :ref:`supported base DataModules
<mednet.libs.classification.config.datamodules>`.
André Anjos
committed
.. autosummary::
André Anjos
committed
:toctree: api/config.datamodule-instances
André Anjos
committed
:template: config.rst
mednet.libs.classification.config.data.indian.default
mednet.libs.classification.config.data.montgomery.default
mednet.libs.classification.config.data.montgomery_shenzhen.default
mednet.libs.classification.config.data.montgomery_shenzhen_indian.default
mednet.libs.classification.config.data.montgomery_shenzhen_indian_padchest.default
mednet.libs.classification.config.data.montgomery_shenzhen_indian_tbx11k.v1_healthy_vs_atb
mednet.libs.classification.config.data.montgomery_shenzhen_indian_tbx11k.v2_others_vs_atb
mednet.libs.classification.config.data.nih_cxr14.default
mednet.libs.classification.config.data.nih_cxr14_padchest.idiap
mednet.libs.classification.config.data.padchest.idiap
mednet.libs.classification.config.data.shenzhen.default
mednet.libs.classification.config.data.tbx11k.v1_healthy_vs_atb
mednet.libs.classification.config.data.tbx11k.v2_others_vs_atb
mednet.libs.classification.config.data.tbx11k.v2_others_vs_atb
André Anjos
committed
.. _mednet.libs.classification.config.datamodule-instances.folds:
André Anjos
committed
Cross-validation DataModules
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
We support cross-validation with precise preset folds. In this section, you
will find the configuration for the first fold (fold-0) for all supported
DataModules. Nine other folds are available for every configuration (from 1 to
9), making up 10 folds per supported DataModule.
André Anjos
committed
:toctree: api/config.datamodule-folds
mednet.libs.classification.config.data.hivtb.fold_0
mednet.libs.classification.config.data.indian.fold_0
mednet.libs.classification.config.data.montgomery.fold_0
mednet.libs.classification.config.data.montgomery_shenzhen.fold_0
mednet.libs.classification.config.data.montgomery_shenzhen_indian.fold_0
mednet.libs.classification.config.data.montgomery_shenzhen_indian_tbx11k.v1_fold_0
mednet.libs.classification.config.data.montgomery_shenzhen_indian_tbx11k.v2_fold_0
mednet.libs.classification.config.data.shenzhen.fold_0
mednet.libs.classification.config.data.tbpoc.fold_0
mednet.libs.classification.config.data.tbx11k.v1_fold_0
mednet.libs.classification.config.data.tbx11k.v2_fold_0
André Anjos
committed
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
.. _mednet.libs.segmentation.config:
------------------------------------
Classification Preset Configurations
------------------------------------
This module contains preset configurations for baseline CNN architectures and
DataModules in a segmentation task.
.. _mednet.libs.segmentation.config.models:
Pre-configured Models
^^^^^^^^^^^^^^^^^^^^^
Pre-configured models you can readily use.
.. autosummary::
:toctree: api/config.models
:template: config.rst
mednet.libs.segmentation.config.models.driu_bn
mednet.libs.segmentation.config.models.driu_od
mednet.libs.segmentation.config.models.driu_pix
mednet.libs.segmentation.config.models.driu
mednet.libs.segmentation.config.models.hed
mednet.libs.segmentation.config.models.lwnet
mednet.libs.segmentation.config.models.m2unet
mednet.libs.segmentation.config.models.unet
.. _mednet.libs.segmentation.config.datamodules:
DataModule support
^^^^^^^^^^^^^^^^^^
Base DataModules and raw data loaders for the various databases currently
supported in this package, for your reference. Each pre-configured DataModule
can receive the name of one or more splits as argument to build a fully
functional DataModule that can be used in training, prediction or testing.
.. autosummary::
:toctree: api/config.datamodules
mednet.libs.segmentation.config.data.chasedb1.datamodule
mednet.libs.segmentation.config.data.cxr8.datamodule
mednet.libs.segmentation.config.data.drhagis.datamodule
mednet.libs.segmentation.config.data.drionsdb.datamodule
mednet.libs.segmentation.config.data.drishtigs1.datamodule
mednet.libs.segmentation.config.data.drive.datamodule
mednet.libs.segmentation.config.data.hrf.datamodule
mednet.libs.segmentation.config.data.iostar.datamodule
mednet.libs.segmentation.config.data.jsrt.datamodule
mednet.libs.segmentation.config.data.montgomery.datamodule
mednet.libs.segmentation.config.data.refuge.datamodule
mednet.libs.segmentation.config.data.rimoner3.datamodule
mednet.libs.segmentation.config.data.shenzhen.datamodule
mednet.libs.segmentation.config.data.stare.datamodule
.. _mednet.libs.segmentation.config.datamodule-instances:
Pre-configured DataModules
^^^^^^^^^^^^^^^^^^^^^^^^^^
DataModules provide access to preset pytorch dataloaders for training,
validating, testing and running prediction tasks. Each of the pre-configured
DataModule is based on one (or more) of the :ref:`supported base DataModules
<mednet.libs.segmentation.config.datamodules>`.
.. autosummary::
:toctree: api/config.datamodule-instances
:template: config.rst
mednet.libs.segmentation.config.data.chasedb1.first_annotator
mednet.libs.segmentation.config.data.chasedb1.second_annotator
mednet.libs.segmentation.config.data.cxr8.default
mednet.libs.segmentation.config.data.drhagis.default
mednet.libs.segmentation.config.data.drionsdb.expert1
mednet.libs.segmentation.config.data.drionsdb.expert2
mednet.libs.segmentation.config.data.drishtigs1.optic_cup_all
mednet.libs.segmentation.config.data.drishtigs1.optic_cup_any
mednet.libs.segmentation.config.data.drishtigs1.optic_disc_all
mednet.libs.segmentation.config.data.drishtigs1.optic_disc_any
mednet.libs.segmentation.config.data.drive.default
mednet.libs.segmentation.config.data.drive.drive_2nd
mednet.libs.segmentation.config.data.hrf.default
mednet.libs.segmentation.config.data.iostar.optic_disc
mednet.libs.segmentation.config.data.iostar.vessel
mednet.libs.segmentation.config.data.jsrt.default
mednet.libs.segmentation.config.data.montgomery.default
mednet.libs.segmentation.config.data.refuge.disc
mednet.libs.segmentation.config.data.refuge.cup
mednet.libs.segmentation.config.data.rimoner3.cup_exp1
mednet.libs.segmentation.config.data.rimoner3.cup_exp2
mednet.libs.segmentation.config.data.rimoner3.disc_exp1
mednet.libs.segmentation.config.data.rimoner3.disc_exp2
mednet.libs.segmentation.config.data.shenzhen.default
mednet.libs.segmentation.config.data.stare.ah
mednet.libs.segmentation.config.data.stare.vk
------------------
Data Augmentations
------------------
Sequences of data augmentations you can readily use.
.. _mednet.libs.common.config.augmentations:
.. autosummary::
:toctree: api/config.augmentations
:template: config.rst
mednet.libs.common.config.augmentations.elastic
mednet.libs.common.config.augmentations.affine