Table of Contents

Interface ISelfSupervisedLoss<T>

Namespace
AiDotNet.Interfaces
Assembly
AiDotNet.dll

Interface for self-supervised loss functions used in meta-learning.

public interface ISelfSupervisedLoss<T>

Type Parameters

T

The numeric type used for calculations (e.g., float, double).

Remarks

Self-supervised learning creates artificial tasks from unlabeled data, allowing models to learn useful representations without explicit labels. This is particularly valuable in meta-learning where the query set is often large but unlabeled.

For Beginners: Self-supervised learning is like learning by creating your own practice problems.

Example: Rotation prediction for images

  • Take an unlabeled image
  • Rotate it by 0°, 90°, 180°, or 270°
  • Train the model to predict which rotation was applied
  • The model learns spatial relationships and features without needing class labels

This is powerful because:

  1. You can use unlabeled data (which is often abundant)
  2. The model learns useful features automatically
  3. These features help with the actual task (classification, etc.)

Think of it like learning to recognize faces by first learning to identify if a photo is upside down. You don't need to know who the person is to learn about facial features!

Common Self-Supervised Tasks: - Rotation Prediction: Predict rotation angle (0°, 90°, 180°, 270°) - Jigsaw Puzzles: Solve scrambled image patches - Colorization: Predict color from grayscale - Context Prediction: Predict spatial relationships between patches - Contrastive Learning: Learn to distinguish similar vs dissimilar examples

Methods

CreateTask<TInput, TOutput>(TInput)

Creates a self-supervised task from unlabeled input data.

(TInput augmentedX, TOutput augmentedY) CreateTask<TInput, TOutput>(TInput input)

Parameters

input TInput

Unlabeled input data (e.g., images).

Returns

(TInput Input, TOutput Output)

A tuple containing:

  • augmentedX: Transformed input data for the self-supervised task
  • augmentedY: Labels for the self-supervised task (e.g., rotation angles)

Type Parameters

TInput
TOutput

Remarks

This method transforms unlabeled data into a supervised learning problem by creating artificial labels based on the transformation applied.

For Beginners: This converts "unlabeled data" into a "labeled learning problem".

Example for rotation prediction:

  • Input: 10 unlabeled images
  • Output: 40 labeled images (each original rotated 4 times: 0°, 90°, 180°, 270°)
  • Labels: [0, 1, 2, 3] indicating which rotation was applied

The model learns to recognize rotations, which teaches it about:

  • Edge orientations
  • Spatial relationships
  • Object structure

These learned features are useful for the actual classification task!

Edit this page