Let All be Whitened: Multi-teacher Distillation for Efficient Visual Retrieval

TL;DR

This paper introduces Whiten-MTD, a multi-teacher distillation framework that improves the efficiency of visual retrieval models by transferring knowledge from pre-trained models to lightweight students through a whitening process.

Contribution

The paper proposes a novel whitening technique for multi-teacher distillation in visual retrieval, enabling effective knowledge transfer from diverse models to lightweight retrieval systems.

Findings

Effective on multiple datasets for image and video retrieval

Balances retrieval accuracy with computational efficiency

Outperforms baseline methods in experiments

Abstract

Visual retrieval aims to search for the most relevant visual items, e.g., images and videos, from a candidate gallery with a given query item. Accuracy and efficiency are two competing objectives in retrieval tasks. Instead of crafting a new method pursuing further improvement on accuracy, in this paper we propose a multi-teacher distillation framework Whiten-MTD, which is able to transfer knowledge from off-the-shelf pre-trained retrieval models to a lightweight student model for efficient visual retrieval. Furthermore, we discover that the similarities obtained by different retrieval models are diversified and incommensurable, which makes it challenging to jointly distill knowledge from multiple models. Therefore, we propose to whiten the output of teacher models before fusion, which enables effective multi-teacher distillation for retrieval models. Whiten-MTD is conceptually simple and practically effective. Extensive experiments on two landmark image retrieval datasets and one video retrieval dataset demonstrate the effectiveness of our proposed method, and its good balance of retrieval performance and efficiency. Our source code is released at https://github.com/Maryeon/whiten_mtd.