Activation Functions
Maxout
Introduced by Goodfellow et al. in Maxout NetworksThe Maxout Unit is a generalization of the ReLU and the leaky ReLU functions. It is a piecewise linear function that returns the maximum of the inputs, designed to be used in conjunction with dropout. Both ReLU and leaky ReLU are special cases of Maxout.
$$f\left(x\right) = \max\left(w^{T}_{1}x + b_{1}, w^{T}_{2}x + b_{2}\right)$$
The main drawback of Maxout is that it is computationally expensive as it doubles the number of parameters for each neuron.
Source: Maxout Networks
Papers
| Paper | Code | Results | Date | Stars |
|---|
Tasks
| Task | Papers | Share |
|---|---|---|
| Image Classification | 9 | 15.00% |
| General Classification | 8 | 13.33% |
| Speech Recognition | 4 | 6.67% |
| Object Detection | 3 | 5.00% |
| Semantic Segmentation | 3 | 5.00% |
| Image Generation | 2 | 3.33% |
| Graph Classification | 2 | 3.33% |
| Object Recognition | 2 | 3.33% |
| Face Verification | 2 | 3.33% |
Usage Over Time
Components
| Component | Type |
|
|---|---|---|
| 🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |
