Semantic Autoencoder for Zero-Shot Learning，Elyor Kodirov Tao Xiang Shaogang Gong，Queen Mary University of London, UK，{e.kodirov, t.xiang, s.gong}@qmul.ac.uk
亮点
- 通过对耦学习提升零次学习系统的性能（类似CycleGan）
- 结构非常简洁，且可直接求解，速度非常快
- 有效应用到其他相关任务（监督聚类）上，证明了范化性能
方法
Linear autoencoder
Model Formulation
which is a well-known Sylvester equation which can be solved efficiently by the Bartels-Stewart algorithm (matlab sylvester).
零次学习：基于以上算法有两种测试的方法：
- 将一个未知的类别特征样本xi通过W映射到语义空间（属性）si，通过比较语义空间的距离找到离它最近的类别（无训练样本），即为它的标签
- 将所有无训练数据类别的语义特征S通过WT映射到特征空间X，通过比较一个未知类别的样本xi和映射到特征空间的类别中心X的距离，找到离它最近的类别，即为它的标签
- 以上两种算法得到结果的准确度基本相同。
监督聚类：在这个问题中，语义空间即为类别标签空间（one-hot class label）。所有测试数据被影射到训练类别标签空间，然后使用k-means聚合
与已有模型的关系：零度学习已有模型一般学习一个满足以下条件的影射：
或者，在［54］中将属性影射到特征空间，学习目标变为，
文中的算法结合了这两者，而且由于W*=WT，在对耦学习中W不可能太大（否则，x乘以两个范数很大的的矩阵无法恢复原来的初始值），正则化项可以被忽略。
实验
零次学习
数据集：Semantic word vector representation is used for large-scale datasets (ImNet-1 and ImNet-2). We train a skip-gram text model on a corpus of 4.6M Wikipedia documents to obtain the word2vec2 [38, 37] word vectors.
特征：除 ImNet-1用AlexNet提取外，其他均使用了GoogleNet
结果：
- Our SAE model achieves the best results on all 6 datasets.
- On the smallscale datasets, the gap between our model’s results to the strongest competitor ranges from 3.5% to 6.5%.
- On the large-scale datasets, the gaps are even bigger: On the largest ImNet-2, our model improves over the state-of-the-art SS-Voc [22] by 8.8%.
- Both the encoder and decoder projection functions in our SAE model (SAE (W) and SAE (WT) respectively) can be used for effective ZSL.
- The encoder projection function seems to be slightly better overall.
- Measures how well a zero-shot learning method can trade-off between recognising data from seen classes and that of unseen classes
- Holding out 20% of the data samples from the seen classes and mixing them with the samples from the unseen classes.
- On AwA, our model is slightly worse than the SynCstruct [13].
- However, on the more challenging CUB dataset, our method significantly outperforms the competitors.
聚类
数据集： A synthetic dataset and Oxford Flowers-17 (848 images)
结果：
- On computational cost, our model (93s) is more expensive than MLCA (39%) but much better than all others (hours~days).
- Achieves the best clustering accuracy