CVAILGMLApr 9, 2019

Few-Shot Learning with Localization in Realistic Settings

arXiv:1904.08502v2174 citations
Originality Incremental advance
AI Analysis

This addresses the gap between artificial few-shot benchmarks and real-world recognition problems, offering incremental improvements for computer vision applications.

The paper tackled the problem of few-shot learning in realistic settings with heavy-tailed class distributions and cluttered scenes, showing that prior methods fail and introducing improvements that double accuracy on a new benchmark.

Traditional recognition methods typically require large, artificially-balanced training classes, while few-shot learning methods are tested on artificially small ones. In contrast to both extremes, real world recognition problems exhibit heavy-tailed class distributions, with cluttered scenes and a mix of coarse and fine-grained class distinctions. We show that prior methods designed for few-shot learning do not work out of the box in these challenging conditions, based on a new "meta-iNat" benchmark. We introduce three parameter-free improvements: (a) better training procedures based on adapting cross-validation to meta-learning, (b) novel architectures that localize objects using limited bounding box annotations before classification, and (c) simple parameter-free expansions of the feature space based on bilinear pooling. Together, these improvements double the accuracy of state-of-the-art models on meta-iNat while generalizing to prior benchmarks, complex neural architectures, and settings with substantial domain shift.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes