1 LearnableFourierFeatures

vin.pose_encoding.LearnableFourierFeatures(config)

Learnable Fourier Features (LFF) positional encoding.

This module maps continuous inputs \(x \in \mathbb{R}^D\) into a learned feature space by:

learning a projection matrix \(W_r\),
applying sinusoidal features, and
mapping them through a small MLP.

Compared to fixed random Fourier features, the learned projection and the MLP allow the encoding to adapt to the downstream task.

1.1 Methods

Name	Description
forward	TODO: Ensure same signature as other pose encoders - i.e. ShellShPoseEncoder.forward (derive both from a common base class!)

1.1.1 forward

vin.pose_encoding.LearnableFourierFeatures.forward(x)

TODO: Ensure same signature as other pose encoders - i.e. ShellShPoseEncoder.forward (derive both from a common base class!)

# LearnableFourierFeatures { #aria_nbv.vin.pose_encoding.LearnableFourierFeatures }

```python
vin.pose_encoding.LearnableFourierFeatures(config)
```

Learnable Fourier Features (LFF) positional encoding.

This module maps continuous inputs $x \in \mathbb{R}^D$ into a learned
feature space by:

1) learning a projection matrix $W_r$,
2) applying sinusoidal features, and
3) mapping them through a small MLP.

Compared to fixed random Fourier features, the learned projection and the
MLP allow the encoding to adapt to the downstream task.

## Methods

| Name | Description |
| --- | --- |
| [forward](#aria_nbv.vin.pose_encoding.LearnableFourierFeatures.forward) | TODO: Ensure same signature as other pose encoders - i.e. ShellShPoseEncoder.forward (derive both from a common base class!) |

### forward { #aria_nbv.vin.pose_encoding.LearnableFourierFeatures.forward }

```python
vin.pose_encoding.LearnableFourierFeatures.forward(x)
```

TODO: Ensure same signature as other pose encoders - i.e. ShellShPoseEncoder.forward (derive both from a common base class!)