ENet Bottleneck
Introduced by Paszke et al. in ENet: A Deep Neural Network Architecture for Real-Time Semantic SegmentationENet Bottleneck is an image model block used in the ENet semantic segmentation architecture. Each block consists of three convolutional layers: a 1 × 1 projection that reduces the dimensionality, a main convolutional layer, and a 1 × 1 expansion. We place Batch Normalization and PReLU between all convolutions. If the bottleneck is downsampling, a max pooling layer is added to the main branch. Also, the first 1 × 1 projection is replaced with a 2 × 2 convolution with stride 2 in both dimensions. We zero pad the activations, to match the number of feature maps.
Source: ENet: A Deep Neural Network Architecture for Real-Time Semantic Segmentation
Latest Papers
Tasks
| TASK | PAPERS | SHARE |
|---|---|---|
| Semantic Segmentation | 10 | 31.25% |
| Autonomous Driving | 4 | 12.50% |
| Instance Segmentation | 2 | 6.25% |
| Scene Understanding | 2 | 6.25% |
| Food Recognition | 1 | 3.13% |
| Image Classification | 1 | 3.13% |
| Self-Driving Cars | 1 | 3.13% |
| Autonomous Vehicles | 1 | 3.13% |
| Knowledge Distillation | 1 | 3.13% |
Usage Over Time
Components
| COMPONENT | TYPE |
|
|---|---|---|
1x1 Convolution
|
Convolutions | |
Batch Normalization
|
Normalization | |
Convolution
|
Convolutions | |
Max Pooling
|
Pooling Operations | |
PReLU
|
Activation Functions | |
SpatialDropout
|
Regularization |