Adaptive Distillation: Aggregating Knowledge from Multiple Paths for Efficient Distillation

TL;DR

This paper introduces an adaptive method for aggregating multiple knowledge distillation paths, improving the generalization of student models across various tasks by dynamically adjusting path importance.

Contribution

It proposes a novel adaptive aggregation approach based on multitask learning to optimize knowledge distillation from multiple paths.

Findings

The adaptive approach outperforms baseline methods in classification.

It improves performance in semantic segmentation tasks.

It enhances object detection accuracy.

Abstract

Knowledge Distillation is becoming one of the primary trends among neural network compression algorithms to improve the generalization performance of a smaller student model with guidance from a larger teacher model. This momentous rise in applications of knowledge distillation is accompanied by the introduction of numerous algorithms for distilling the knowledge such as soft targets and hint layers. Despite this advancement in different techniques for distilling the knowledge, the aggregation of different paths for distillation has not been studied comprehensively. This is of particular significance, not only because different paths have different importance, but also due to the fact that some paths might have negative effects on the generalization performance of the student model. Hence, we need to adaptively adjust the importance of each path to maximize the impact of distillation on the student model. In this paper, we explore different approaches for aggregating these different paths and introduce our proposed adaptive approach based on multitask learning methods. We empirically demonstrate the effectiveness of the proposed approach over other baselines on the applications of knowledge distillation in classification, semantic segmentation, and object detection tasks.