Zero-Shot Learning via Joint Latent Similarity Embedding

Abstract

Zero-shot recognition (ZSR) deals with the problem of predicting class labels for target domain instances based on source domain side information (e.g. attributes) of unseen classes. We formulate ZSR as a binary prediction problem. Our resulting classifier is class-independent. It takes an arbitrary pair of source and target domain instances as input and predicts whether or not they come from the same class, i.e. whether there is a match. We model the posterior probability of a match since it is a sufficient statistic and propose a latent probabilistic model in this context. We develop a joint discriminative learning framework based on dictionary learning to jointly learn the parameters of our model for both domains, which ultimately leads to our class-independent classifier. Many of the existing embedding methods can be viewed as special cases of our probabilistic model. On ZSR our method shows 4.90% improvement over the state-of-the-art in accuracy averaged across four benchmark datasets. We also adapt ZSR method for zero-shot retrieval and show 22.45% improvement accordingly in mean average precision (mAP).

Describing Unseen Classes by Exemplars: Zero-Shot Learning Using Grouped Simile Ensemble

“…Discussion Our method can outperform most of the stateof-the-art methods and the overall recognition rate is only 0.7 % lower than that of [41] on AwA. However, our method achieves significant improvement of 6.03% over [41] on the aPY dataset. We ascribe such performance difference to that the variation of unseen classes of the two datasets is different.…”

“…Many recent approaches adopt such an embedding manner and achieve promising results [13,4,33,15,7,19,39,8,23]. Besides, similarity-based frameworks also adopt the embedding approach [24,40,41,34,8,25]. But the semantic space aims to associate unseen to seen classes.…”

“…Another mainstream of semantic representations is similarity-based. Class-wise similarities can be obtained by either human annotators [38,20] or based on the textual descriptions [40,41]. Our simile description also shares the idea of similarity comparison.…”

Towards Fine-Grained Open Zero-Shot Learning: Inferring Unseen Visual Features from Attributes

“…Moreover, such a framework be straightforwardly combined with Deep Neraul Network [26]. The much recent research adopts the embedding approach and demonstrates state-of-the-art performance [30,2,39,40,11,25]. The remaining challenges so far is to break the restrictions of conventional ZSL settings.…”

“…ESEZL [30] combines visualattribute and attribute-label embedding into one joint func- tion. SSE [39] and JLSE [40] are similarity-based approaches that jointly learn a dictionary learning function for both visual and attribute domains. Note that all of the compared methods use attributes as side information.…”