Model Architectures¶

Deep-neural network models are categorized by tasks.

Classification¶

Pre-configured models supporting classification tasks.

Table 1 Pre-configured classification models.¶
Config. key	Module	Base type
`alexnet`	`config.classify.models.alexnet`	`models.classify.alexnet.Alexnet`
`alexnet-pretrained`	`config.classify.models.alexnet_pretrained`	`models.classify.alexnet.Alexnet`
`cnn3d`	`config.classify.models.cnn3d`	`models.classify.cnn3d.Conv3DNet`
`densenet`	`config.classify.models.densenet`	`models.classify.densenet.Densenet`
`densenet-pretrained`	`config.classify.models.densenet_pretrained`	`models.classify.densenet.Densenet`
`logistic-regression`	`config.classify.models.logistic_regression`	`models.classify.logistic_regression.LogisticRegression`
`mlp`	`config.classify.models.mlp`	`models.classify.mlp.MultiLayerPerceptron`
`pasa`	`config.classify.models.pasa`	`models.classify.pasa.Pasa`

Semantic Segmentation¶

Pre-configured models supporting semantic segmentation tasks.

Table 2 Pre-configured semantic segmentation models¶
Config. key	Module	Base type
`driu`	`config.segment.models.driu`	`models.segment.driu.DRIU`
`driu-bn`	`config.segment.models.driu_bn`	`models.segment.driu_bn.DRIUBN`
`driu-od`	`config.segment.models.driu_od`	`models.segment.driu_od.DRIUOD`
`driu-pix`	`config.segment.models.driu_pix`	`models.segment.driu_pix.DRIUPix`
`hed`	`config.segment.models.hed`	`models.segment.hed.HED`
`lwnet`	`config.segment.models.lwnet`	`models.segment.lwnet.LittleWNet`
`m2unet`	`config.segment.models.m2unet`	`models.segment.m2unet.M2Unet`
`unet`	`config.segment.models.unet`	`models.segment.unet.Unet`

Object Detection¶

Pre-configured models supporting object detection tasks.

Table 3 Pre-configured object detection models¶
Config. key	Module	Base type
`faster-rcnn`	`config.detect.models.faster_rcnn`	`models.detect.faster_rcnn.FasterRCNN`