Negative Margin Matters: Understanding Margin in Few-shot Classification

This paper introduces a negative margin loss to metric learning based few-shot learning methods. The negative margin loss significantly outperforms regular softmax loss, and achieves state-of-the-art accuracy on three standard few-shot classification benchmarks with few bells and whistles. These results are contrary to the common practice in the metric learning field, that the margin is zero or positive. To understand why the negative margin loss performs well for the few-shot classification, we analyze the discriminability of learned features w.r.t different margins for training and novel classes, both empirically and theoretically. We find that although negative margin reduces the feature discriminability for training classes, it may also avoid falsely mapping samples of the same novel class to multiple peaks or clusters, and thus benefit the discrimination of novel classes. Code is available at

PDF Abstract ECCV 2020 PDF ECCV 2020 Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
Few-Shot Image Classification CUB 200 5-way 1-shot Neg-Margin Accuracy 72.66 # 20
Few-Shot Image Classification CUB 200 5-way 5-shot Neg-Margin Accuracy 89.40 # 16
Few-Shot Image Classification Mini-ImageNet - 1-Shot Learning Neg-Margin Accuracy 63.85 # 11
Few-Shot Image Classification Mini-ImageNet to CUB - 5 shot learning Neg-Margin Accuracy 69.30 # 2