BloomCoreset: Fast Coreset Sampling using Bloom Filters for Fine-Grained Self-Supervised Learning

TL;DR

BloomCoreset introduces a Bloom filter-based sampling method that accelerates coreset selection in fine-grained self-supervised learning, maintaining high accuracy while drastically reducing sampling time.

Contribution

The paper presents a novel Bloom filter-based approach for fast coreset sampling in fine-grained SSL, significantly improving efficiency over existing methods.

Findings

Achieves 98.5% reduction in sampling time

Maintains only 0.83% accuracy trade-off

Outperforms baseline sampling strategies

Abstract

The success of deep learning in supervised fine-grained recognition for domain-specific tasks relies heavily on expert annotations. The Open-Set for fine-grained Self-Supervised Learning (SSL) problem aims to enhance performance on downstream tasks by strategically sampling a subset of images (the Core-Set) from a large pool of unlabeled data (the Open-Set). In this paper, we propose a novel method, BloomCoreset, that significantly reduces sampling time from Open-Set while preserving the quality of samples in the coreset. To achieve this, we utilize Bloom filters as an innovative hashing mechanism to store both low- and high-level features of the fine-grained dataset, as captured by Open-CLIP, in a space-efficient manner that enables rapid retrieval of the coreset from the Open-Set. To show the effectiveness of the sampled coreset, we integrate the proposed method into the state-of-the-art fine-grained SSL framework, SimCore [1]. The proposed algorithm drastically outperforms the sampling strategy of the baseline in SimCore [1] with a $98.5\%$ reduction in sampling time with a mere $0.83\%$ average trade-off in accuracy calculated across $11$ downstream datasets.