OmnImage: Evolving 1k Image Cliques for Few-Shot Learning
OmnImage: Evolving 1k Image Cliques for Few-Shot Learning
复制标题
DOI：
10.1145/3583131.3590430
复制
发表时间：
2023-07
期刊：
影响因子：
--
通讯作者：
Lapo Frati;Neil Traft;Nick Cheney
中科院分区：
文献类型：
--
作者：
Lapo Frati;Neil Traft;Nick Cheney
少镜头学习数据集包含大量的类，每个类中只有几个例子。现有的数据集可能包含数千个类，但非常简单的图像（例如手写字符），因此简单的基线可以表现得非常好。或者它们可能非常复杂，类内变化很大，背景分散，以至于它们太难了，无法用这么少的例子进行有意义的学习。为了构建一致的自然图像的可定制数据集，我们组装了一个新的数据集，每个类包含ImageNet-1 k的1000个类中的每个类的一个小子集。为了选择具有明确类内一致性的子集，我们使用进化方法来最小化由预训练的VGG模型生成的特征之间的成对余弦距离。我们在这些进化的图像集团上训练分类器，发现我们的进化数据集比手写数字提供了更大的挑战，但不是ImageNet的非进化子集的极端困难。我们发现，在这些进化的原型类上预训练我们的分类器显著提高了对ImageNet随机子集进行分类的性能（相对于在类似的随机类子集上进行预训练），并推测这些原型类可能有利于播种概念学习。数据集和代码可在https://github.com/lfrati/OmnImage上公开获取
Few-shot learning datasets contain a large number of classes with only a few examples in each. Existing datasets may contain thousands of classes, but very simple images (e.g. handwritten characters) such that a naive baseline can perform very well. Or they may be so complex, with large within-class variation and distracting background, that they are too difficult to enable meaningful learning with so few examples. To construct a customizable dataset of consistent natural images, we assemble a new dataset with each class containing a small subset from each of the 1000 classes of ImageNet-1k. To select subsets with clear within-class consistency we use an evolutionary approach to minimize the pairwise cosine-distance between features generated by a pre-trained VGG model. We train a classifier on these evolved image cliques and find that our evolved dataset provides a greater challenge than hand written digits, but not the extreme difficulty of a non-evolved subset of ImageNet. We find that pre-training our classifier on these evolved prototypical classes significantly improves performance on classifying random subsets ImageNet (relative to pre-training on similar random-class subsets), and conjecture that these prototypical classes may be beneficial for seeding concept learning. Dataset and code is publicly available at: https://github.com/lfrati/OmnImage