Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Label Distribution Learning from Logical Label

Yuheng Jia, Jiawei Tang, Jiahao Jiang

TL;DR

This work proposes a novel method to learn an LDL model directly from the logical label, which unifies LE and LDL into a joint model, and avoids the drawbacks of the previous LE methods.

Abstract

Label distribution learning (LDL) is an effective method to predict the label description degree (a.k.a. label distribution) of a sample. However, annotating label distribution (LD) for training samples is extremely costly. So recent studies often first use label enhancement (LE) to generate the estimated label distribution from the logical label and then apply external LDL algorithms on the recovered label distribution to predict the label distribution for unseen samples. But this step-wise manner overlooks the possible connections between LE and LDL. Moreover, the existing LE approaches may assign some description degrees to invalid labels. To solve the above problems, we propose a novel method to learn an LDL model directly from the logical label, which unifies LE and LDL into a joint model, and avoids the drawbacks of the previous LE methods. Extensive experiments on various datasets prove that the proposed approach can construct a reliable LDL model directly from the logical label, and produce more accurate label distribution than the state-of-the-art LE methods.

Label Distribution Learning from Logical Label

TL;DR

This work proposes a novel method to learn an LDL model directly from the logical label, which unifies LE and LDL into a joint model, and avoids the drawbacks of the previous LE methods.

Abstract

Label distribution learning (LDL) is an effective method to predict the label description degree (a.k.a. label distribution) of a sample. However, annotating label distribution (LD) for training samples is extremely costly. So recent studies often first use label enhancement (LE) to generate the estimated label distribution from the logical label and then apply external LDL algorithms on the recovered label distribution to predict the label distribution for unseen samples. But this step-wise manner overlooks the possible connections between LE and LDL. Moreover, the existing LE approaches may assign some description degrees to invalid labels. To solve the above problems, we propose a novel method to learn an LDL model directly from the logical label, which unifies LE and LDL into a joint model, and avoids the drawbacks of the previous LE methods. Extensive experiments on various datasets prove that the proposed approach can construct a reliable LDL model directly from the logical label, and produce more accurate label distribution than the state-of-the-art LE methods.
Paper Structure (17 sections, 22 equations, 2 figures, 4 tables, 2 algorithms)

This paper contains 17 sections, 22 equations, 2 figures, 4 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Works
  3. Label Distribution Learning
  4. Label Enhancement
  5. The Proposed Approach
  6. Optimization
  7. Generalization Bound
  8. Experiments
  9. Baselines and Settings
  10. Visualization of the Recovered Label Distributions
  11. Results
  12. Recovery Results
  13. Predictive Results
  14. Further Analysis
  15. Ablation Study
  16. ...and 2 more sections

Figures (2)

  • Figure 1: An example of using label distribution to describe a natural scene image in (a). The histogram in (b) denotes the logical label, indicating whether a label can describe the image in (a), and the line chart in (c) shows the label distribution, revealing to what degree a label can describe the image in (a).
  • Figure 2: The visualization of two typical recovery results on the Twitter (left) and NS (right) dataset.