Δ
\Delta
Δ-ENCODER:一种有效的样本合成方法,用于小样本识别
Delta Encoder: An Effective Sample Synthesis Method for Few Shot Object Recognition
Delta Encoder: An Effective Sample Synthesis Method for Few Shot Object Recognition
Abstract
Δ \Delta Δ-ENCODER:学习仅基于一个或几个示例对新类别进行分类是现代计算机视觉中的一个长期挑战。在这项工作中,我们提出了一种简单而有效的方法,用于少样本(和单样本)物体识别。我们的方法基于一个改进的自动编码器,表示为 Δ \Delta Δ-encoder ,可以通过查看来自其中的几个示例来学习合成未见类别的新的样本。然后将合成的样本用于训练分类器。** 所提出的方法学习提取同类训练实例对之间的可迁移的类内变形或“变化”,并将这些变化应用于少数提供的新类别(训练阶段未见)的例子,以便有效地合成新类别的样本。** 所提出的方法改善了单样本物体识别中的最新技术,并且在少数情况下比较有利。一经接受,代码将被提供。
Learning to classify new categories based on just one or a few examples is a long-standing challenge in modern computer vision. In this work, we proposes a simple yet effective method for few-shot (and one-shot) object recognition. Our approach is based on a modified auto-encoder, denoted Δ \Delta Δ-encoder, that learns to synthesize new samples for an unseen category just by seeing few examples from it. The synthesized samples are then used to train a classifier. The proposed approach learns to both extract transferable intra-class deformations, or “deltas”, between same-class pairs of training examples, and to apply those deltas to the few provided examples of a novel class (unseen during training) in order to efficiently synthesize samples from that new class. The proposed method improves over the state-of-the-art in one-shot object-recognition and compares favorably in the few-shot case. Upon acceptance code will be made available.
问题
小样本问题。具体来说是解决样本不足的问题。
方法
一种基于自编码器的样本生成方法。具体来说,提取训练类别的实例对之间的变化/差异,将该差异应用在仅有的新类别的少数样本上,生成新的样本。
数据
输入:2048维的VGG或Resnet特征
输出:1024个新类的样本