BECLR: Batch Enhanced Contrastive Few-Shot Learning
Learning quickly from very few labeled samples is a fundamental attribute that separates machines and humans in the era of deep representation learning. Unsupervised few-shot learning (U-FSL) aspires to bridge this gap by discarding the reliance on annotations at training time. Intrigued by the success of contrastive learning approaches in the realm of U-FSL, we structurally approach their shortcomings in both pretraining and downstream inference stages. We propose a novel Dynamic Clustered mEmory (DyCE) module to promote a highly separable latent representation space for enhancing positive sampling at the pretraining phase and infusing implicit class-level insights into unsupervised contrastive learning. We then tackle the, somehow overlooked yet critical, issue of sample bias at the few-shot inference stage. We propose an iterative Optimal Transport-based distribution Alignment (OpTA) strategy and demonstrate that it efficiently addresses the problem, especially in low-shot scenarios where FSL approaches suffer the most from sample bias. We later on discuss that DyCE and OpTA are two intertwined pieces of a novel end-to-end approach (we coin as BECLR), constructively magnifying each other's impact. We then present a suite of extensive quantitative and qualitative experimentation to corroborate that BECLR sets a new state-of-the-art across ALL existing U-FSL benchmarks (to the best of our knowledge), and significantly outperforms the best of the current baselines (codebase available at: https://github.com/stypoumic/BECLR).
PDF Abstract ICLR 2024 PDFCode
Results from the Paper
Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Benchmark |
---|---|---|---|---|---|---|
Unsupervised Few-Shot Image Classification | Mini-Imagenet 5-way (1-shot) | BECLR | Accuracy | 80.57 | # 1 | |
Unsupervised Few-Shot Image Classification | Mini-Imagenet 5-way (5-shot) | BECLR | Accuracy | 87.82 | # 1 | |
Unsupervised Few-Shot Image Classification | Tiered ImageNet 5-way (1-shot) | BECLR | Accuracy | 81.69 | # 1 | |
Unsupervised Few-Shot Image Classification | Tiered ImageNet 5-way (5-shot) | BECLR | Accuracy | 87.86 | # 1 |