Log In Sign Up

Label-similarity Curriculum Learning

by   Urun Dogan, et al.

Curriculum learning can improve neural network training by guiding the optimization to desirable optima. We propose a novel curriculum learning approach for image classification that adapts the loss function by changing the label representation. The idea is to use a probability distribution over classes as target label, where the class probabilities reflect the similarity to the true class. Gradually, this label representation is shifted towards the standard one-hot-encoding. That is, in the beginning minor mistakes are corrected less than large mistakes, resembling a teaching process in which broad concepts are explained first before subtle differences are taught. The class similarity can be based on prior knowledge. For the special case of the labels being natural words, we propose a generic way to automatically compute the similarities. The natural words are embedded into Euclidean space using a standard word embedding. The probability of each class is then a function of the cosine similarity between the vector representations of the class and the true label. The proposed label-similarity curriculum learning (LCL) approach was empirically evaluated on several popular deep learning architectures for image classification task applied to three datasets, ImageNet, CIFAR100, and AWA2. In all scenarios, LCL was able to improve the classification accuracy on the test data compared to standard training.


Statistical Measures For Defining Curriculum Scoring Function

Curriculum learning is a training strategy that sorts the training examp...

Instance-based Label Smoothing For Better Calibrated Classification Networks

Label smoothing is widely used in deep neural networks for multi-class c...

A Deep Model for Partial Multi-Label Image Classification with Curriculum Based Disambiguation

In this paper, we study the partial multi-label (PML) image classificati...

Hierarchical Class-Based Curriculum Loss

Classification algorithms in machine learning often assume a flat label ...

Learning with Hierarchical Complement Objective

Label hierarchies widely exist in many vision-related problems, ranging ...

Improving Model Training via Self-learned Label Representations

Modern neural network architectures have shown remarkable success in sev...

Curriculum Q-Learning for Visual Vocabulary Acquisition

The structure of curriculum plays a vital role in our learning process, ...