ScreenerNet: Learning Self-Paced Curriculum for Deep Neural Networks

TL;DR

ScreenerNet introduces a learnable self-paced curriculum mechanism for deep neural networks, improving convergence speed and accuracy across vision tasks and reinforcement learning by adaptively weighting training samples.

Contribution

This paper presents ScreenerNet, a novel attachable neural network that learns sample weights in an end-to-end manner, enhancing curriculum learning without bias or memory requirements.

Findings

Achieves faster convergence and higher accuracy than existing curriculum methods.

Effective across multiple datasets including MNIST, CIFAR10, and Pascal VOC2012.

Enhances other curriculum techniques like Prioritized Experience Replay.

Abstract

We propose to learn a curriculum or a syllabus for supervised learning and deep reinforcement learning with deep neural networks by an attachable deep neural network, called ScreenerNet. Specifically, we learn a weight for each sample by jointly training the ScreenerNet and the main network in an end-to-end self-paced fashion. The ScreenerNet neither has sampling bias nor requires to remember the past learning history. We show the networks augmented with the ScreenerNet achieve early convergence with better accuracy than the state-of-the-art curricular learning methods in extensive experiments using three popular vision datasets such as MNIST, CIFAR10 and Pascal VOC2012, and a Cart-pole task using Deep Q-learning. Moreover, the ScreenerNet can extend other curriculum learning methods such as Prioritized Experience Replay (PER) for further accuracy improvement.