A ResNeXt Block is a type of residual block used as part of the ResNeXt CNN architecture. It uses a "split-transform-merge" strategy (branched paths within a single module) similar to an Inception module, i.e. it aggregates a set of transformations. Compared to a Residual Block, it exposes a new dimension, cardinality (size of set of transformations) $C$, as an essential factor in addition to depth and width.
Formally, a set of aggregated transformations can be represented as: $\mathcal{F}(x)=\sum_{i=1}^{C}\mathcal{T}_i(x)$, where $\mathcal{T}_i(x)$ can be an arbitrary function. Analogous to a simple neuron, $\mathcal{T}_i$ should project $x$ into an (optionally low-dimensional) embedding and then transform it.
Source: Aggregated Residual Transformations for Deep Neural NetworksPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Object Detection | 38 | 13.33% |
Image Classification | 28 | 9.82% |
Semantic Segmentation | 23 | 8.07% |
General Classification | 19 | 6.67% |
Instance Segmentation | 15 | 5.26% |
Classification | 8 | 2.81% |
Action Recognition | 7 | 2.46% |
Panoptic Segmentation | 6 | 2.11% |
Real-Time Object Detection | 5 | 1.75% |
Component | Type |
|
---|---|---|
1x1 Convolution
|
Convolutions | |
Batch Normalization
|
Normalization | |
Convolution
|
Convolutions | |
Grouped Convolution
|
Convolutions | |
ReLU
|
Activation Functions | |
Residual Connection
|
Skip Connections |