**DenseNAS-B** is a mobile convolutional neural network discovered through the [DenseNAS](https://paperswithcode.com/method/densenas) [neural architecture search](https://paperswithcode.com/method/neural-architecture-search) method. The basic building block is MBConvs, or inverted bottleneck residuals, from the [MobileNet](https://paperswithcode.com/method/mobilenetv2) architectures.

**DetNet** is a backbone convolutional neural network for object detection. Different from traditional pre-trained models for ImageNet classification, DetNet maintains the spatial resolution of the features even though extra stages are included. DetNet attempts to stay efficient by employing a low complexity dilated bottleneck structure.

DetNet

DetNet: A Backbone network for Object Detection

DenseNAS-B

Densely Connected Search Space for More Flexible Neural Architecture Search

**StyleALAE** is a type of [adversarial latent autoencoder](https://paperswithcode.com/method/alae) that uses a [StyleGAN](https://paperswithcode.com/method/stylegan) based generator. For this the latent space $\mathcal{W}$ plays the same role as the intermediate latent space in [StyleGAN](https://paperswithcode.com/method/stylegan). Therefore, the $G$ network becomes the part of StyleGAN depicted on the right side of the Figure. The left side is a
novel architecture that we designed to be the encoder $E$. The StyleALAE encoder has [Instance Normalization](https://paperswithcode.com/method/instance-normalization) (IN) layers to extract multiscale style information that is combined into a latent code $w$ via a learnable multilinear map.

Source	Densely Connected Search Space for More Flexible Neural Architecture Search
Year	2000
Data Source	Papers with Code archive (CC BY-SA)