AI-KD: Adversarial learning and Implicit regularization for self-Knowledge Distillation

TL;DR

AI-KD introduces an adversarial and regularization-based self-knowledge distillation method that enhances model training by aligning predictive distributions with pre-trained models, leading to improved performance.

Contribution

The paper proposes a novel AI-KD method combining adversarial learning and implicit regularization for self-knowledge distillation, which better aligns predictive distributions.

Findings

Achieves superior performance over state-of-the-art methods.

Effectively distills deterministic and progressive knowledge.

Improves distribution alignment between models.

Abstract

We present a novel adversarial penalized self-knowledge distillation method, named adversarial learning and implicit regularization for self-knowledge distillation (AI-KD), which regularizes the training procedure by adversarial learning and implicit distillations. Our model not only distills the deterministic and progressive knowledge which are from the pre-trained and previous epoch predictive probabilities but also transfers the knowledge of the deterministic predictive distributions using adversarial learning. The motivation is that the self-knowledge distillation methods regularize the predictive probabilities with soft targets, but the exact distributions may be hard to predict. Our method deploys a discriminator to distinguish the distributions between the pre-trained and student models while the student model is trained to fool the discriminator in the trained procedure. Thus, the student model not only can learn the pre-trained model's predictive probabilities but also align the distributions between the pre-trained and student models. We demonstrate the effectiveness of the proposed method with network architectures on multiple datasets and show the proposed method achieves better performance than state-of-the-art methods.