Domain-Specific Embedding Network for Zero-Shot Recognition

@article{Min2019DomainSpecificEN,
  title={Domain-Specific Embedding Network for Zero-Shot Recognition},
  author={Shaobo Min and Hantao Yao and Hongtao Xie and Zhengjun Zha and Yongdong Zhang},
  journal={Proceedings of the 27th ACM International Conference on Multimedia},
  year={2019}
}
Zero-Shot Learning (ZSL) seeks to recognize a sample from either seen or unseen domain by projecting the image data and semantic labels into a joint embedding space. However, most existing methods directly adapt a well-trained projection from one domain to another, thereby ignoring the serious bias problem caused by domain differences. To address this issue, we propose a novel Domain-Specific Embedding Network (DSEN) that can apply specific projections to different domains for unbiased… 

Figures and Tables from this paper

Semantic Enhanced Cross-modal GAN for Zero-shot Learning
TLDR
A Semantic Enhanced Cross- modal GAN (SECM GAN) is proposed by imposing the cross-modal association for improving the semantic and discriminative property of the generated features, which makes reliance on GAN get reduced.
Semantic-guided Reinforced Region Embedding for Generalized Zero-Shot Learning
TLDR
A novel Semantic-guided Reinforced Region Embedding (SR2E) network that can localize important objects in the long-term interests to construct semantic-visual embedding space is proposed.
Domain-Aware Visual Bias Eliminating for Generalized Zero-Shot Learning
TLDR
A novel Domain-aware Visual Bias Eliminating (DVBE) network is proposed that constructs two complementary visual representations, i.e., semantic-free and semantic-aligned, to treat seen and unseen domains separately and outperforms existing methods by averaged 5.7% improvement.
Task-Independent Knowledge Makes for Transferable Representations for Generalized Zero-Shot Learning
TLDR
A novel Dual-Contrastive Embedding Network (DCEN) that simultaneously learns taskspecific and task-independent knowledge via semantic alignment and instance discrimination and which obtains averaged 4.1% improvement on four public benchmarks.
Self-Adaptive Embedding For Few-Shot Classification By Hierarchical Attention
  • X. Wang, Feng Wu, Jie Wang
  • Computer Science
    2020 IEEE International Conference on Multimedia and Expo (ICME)
  • 2020
TLDR
A novel Self-Adaptive Embedding approach (SAE) is proposed that can effectively capture the most discriminative features at the instance level, which significantly improves its performance on downstream classification tasks and significantly outperforms existing state-of-the-art methods.
Dual Progressive Prototype Network for Generalized Zero-Shot Learning
TLDR
Experiments on four benchmarks prove that DPPN effectively alleviates the domain shift problem in GZSL and benefts the semantic-visual alignment and representation transferability.
Generalized Zero-Shot Learning using Generated Proxy Unseen Samples and Entropy Separation
TLDR
This work proposes to use a generative model (GAN) for synthesizing the visual proxy samples while strictly adhering to the standard assumptions of the GZSL, and hypothesizes that such proxy samples can effectively be used to characterize the average entropy of the label distribution of the samples from the unseen classes.
Relation-aware Compositional Zero-shot Learning for Attribute-Object Pair Recognition
TLDR
A novel model for recognizing images with composite attribute-object concepts, notably for composite concepts that are unseen during model training is proposed and a blocking mechanism is proposed that equalizes the information available to the model for both seen and unseen concepts.
Adaptive Bilinear Pooling for Fine-grained Representation Learning
TLDR
This work proposes a generalized feature interaction method, named Adaptive Bilinear Pooling (ABP), which can adaptively infer a suitable pooling strategy for a given sample based on image content, and demonstrates the effectiveness of the proposed method on three widely used benchmarks.
Learning Adaptive Embedding Considering Incremental Class
TLDR
This work proposes a Class-Incremental Learning without Forgetting framework, which aims to learn adaptive embedding for processing novel class detection and model update in a unified framework and designs to regularize classification with decoupled prototype based loss, which can improve the intra-class and inter-class structure significantly.

References

SHOWING 1-10 OF 55 REFERENCES
Unsupervised Domain Adaptation for Zero-Shot Learning
TLDR
A novel ZSL method is proposed based on unsupervised domain adaptation which uses the target domain class labels' projections in the semantic space to regularise the learned target domain projection thus effectively overcoming the projection domain shift problem.
Semantic Autoencoder for Zero-Shot Learning
TLDR
This work presents a novel solution to ZSL based on learning a Semantic AutoEncoder (SAE), which outperforms significantly the existing ZSL models with the additional benefit of lower computational cost and beats the state-of-the-art when the SAE is applied to supervised clustering problem.
Transductive Multi-view Embedding for Zero-Shot Recognition and Annotation
TLDR
This paper proposes a novel framework, transductive multi-view embedding, that rectifies the projection shift between the auxiliary and target domains, exploits the complementarity of multiple semantic representations, achieves state-of-the-art recognition results on image and video benchmark datasets, and enables novel cross-view annotation tasks.
Learning a Deep Embedding Model for Zero-Shot Learning
  • Li Zhang, T. Xiang, S. Gong
  • Computer Science
    2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)
  • 2017
TLDR
This paper proposes to use the visual space as the embedding space instead of embedding into a semantic space or an intermediate space, and argues that in this space, the subsequent nearest neighbour search would suffer much less from the hubness problem and thus become more effective.
Zero-Shot Visual Recognition Using Semantics-Preserving Adversarial Embedding Networks
TLDR
Through adversarial learning of the two subspaces, SP-AEN can transfer the semantics from the reconstructive subspace to the discriminative one, accomplishing the improved zero-shot recognition of unseen classes.
Pseudo Transfer with Marginalized Corrupted Attribute for Zero-shot Learning
TLDR
This work proposes a pseudo transfer mechanism, where the distribution of unseen data is synthesized using semantic embeddings, then the mismatch between the seen distribution and the synthesized unseen distribution is minimized, and successfully overcomes the Z-bias and Z-variance problems.
Transductive Unbiased Embedding for Zero-Shot Learning
TLDR
This paper proposes a straightforward yet effective method named Quasi-Fully Supervised Learning (QFSL) to alleviate the bias problem in Zero-Shot Learning, which outperforms existing state-of-the-art approaches by a huge margin.
Semantically Consistent Regularization for Zero-Shot Recognition
TLDR
The role of semantics in zero-shot learning is considered, and a new convolutional neural network (CNN) framework is proposed, which proposes the use of semantics as constraints for recognition.
Preserving Semantic Relations for Zero-Shot Learning
TLDR
This work proposes to utilize the structure of the space spanned by the attributes using a set of relations to preserve these relations in the embedding space, thereby inducing semanticity to the embedded space for zero-shot learning.
Discriminative Learning of Latent Features for Zero-Shot Recognition
TLDR
This work proposes an end-to-end network that is capable of automatically discovering discriminative regions by a zoom network and learning discrim inative semantic representations in an augmented space introduced for both user-defined and latent attributes.
...
...