Filters








4,168 Hits in 5.3 sec

Transductive Multi-class and Multi-label Zero-shot Learning [article]

Yanwei Fu, Yongxin Yang, Timothy M. Hospedales, Tao Xiang, Shaogang Gong
2015 arXiv   pre-print
Recently, zero-shot learning (ZSL) has received increasing interest.  ...  , and is used to bridge between these domains for knowledge transfer.  ...  We propose a novel framework for multi-label zero-shot learning [9] .  ... 
arXiv:1503.07884v1 fatcat:or4zahtjj5atpoxks5n4dilega

Robust Zero-Shot Cross-Domain Slot Filling with Example Values

Darsh Shah, Raghav Gupta, Amir Fayazi, Dilek Hakkani-Tur
2019 Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics  
Task-oriented dialog systems increasingly rely on deep learning-based slot filling models, usually needing extensive labeled training data for target domains.  ...  Prior zero-shot slot filling models use slot descriptions to learn concepts, but are not robust to misaligned schemas.  ...  We would also like to thank the Deep Dialogue team at Google Research for their support.  ... 
doi:10.18653/v1/p19-1547 dblp:conf/acl/ShahGFH19 fatcat:qrdmwn45srbdneyourkyaa5nje

On Learning Semantic Representations for Million-Scale Free-Hand Sketches [article]

Peng Xu, Yongye Huang, Tongtong Yuan, Tao Xiang, Timothy M. Hospedales, Yi-Zhe Song, Liang Wang
2020 arXiv   pre-print
knowledge for sketch zero-shot domain alignment.  ...  (ii) We propose a deep embedding model for sketch zero-shot recognition, via collecting a large-scale edge-map dataset and proposing to extract a set of semantic vectors from edge-maps as the semantic  ...  Region Embedding Network for zero-shot learning (AREN) [57] .  ... 
arXiv:2007.04101v1 fatcat:cng2cw6r5fg43p5erfisj57tu4

Recent Advances in Zero-shot Recognition [article]

Yanwei Fu, Tao Xiang, Yu-Gang Jiang, Xiangyang Xue, Leonid Sigal, and Shaogang Gong
2017 arXiv   pre-print
One approach to scaling up the recognition is to develop models capable of recognizing unseen categories without any training instances, or zero-shot recognition/ learning.  ...  or when zero-shot recognition is implemented in a real-world setting.  ...  Yanwei Fu is supported by The Program for Professor of Special Appointment (Eastern Scholar) at Shanghai Institutions of Higher Learning.  ... 
arXiv:1710.04837v1 fatcat:u3mp6dgj2rgqrarjm4dcywegmy

Zero-shot Cross-lingual Dialogue Systems with Transferable Latent Variables

Zihan Liu, Jamin Shin, Yan Xu, Genta Indra Winata, Peng Xu, Andrea Madotto, Pascale Fung
2019 Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)  
Hence, we propose a zero-shot adaptation of task-oriented dialogue system to lowresource languages.  ...  detection and slot filling) compared to the current state-of-the-art model in the zero-shot scenario.  ...  Upadhyay et al. (2018) leveraged joint training and cross-lingual embeddings to do zero-shot and almost zero-shot transfer learning in intent prediction and slot filling.  ... 
doi:10.18653/v1/d19-1129 dblp:conf/emnlp/LiuSXWXMF19 fatcat:5l7jx52axzdoxk2csrvxn47mye

Zero-shot Cross-lingual Dialogue Systems with Transferable Latent Variables [article]

Zihan Liu, Jamin Shin, Yan Xu, Genta Indra Winata, Peng Xu, Andrea Madotto, Pascale Fung
2019 arXiv   pre-print
Hence, we propose a zero-shot adaptation of task-oriented dialogue system to low-resource languages.  ...  detection and slot filling) compared to the current state-of-the-art model in the zero-shot scenario.  ...  Upadhyay et al. (2018) leveraged joint training and cross-lingual embeddings to do zero-shot and almost zero-shot transfer learning in intent prediction and slot filling.  ... 
arXiv:1911.04081v1 fatcat:66wloyvprzcwlgc2rovedyz4au

Reconstructing Capsule Networks for Zero-shot Intent Classification

Han Liu, Xiaotong Zhang, Lu Fan, Xuandi Fu, Qimai Li, Xiao-Ming Wu, Albert Y.S. Lam
2019 Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)  
However, it has two unaddressed limitations: (1) it cannot deal with polysemy when extracting semantic capsules; (2) it hardly recognizes the utterances of unseen intents in the generalized zero-shot intent  ...  A recently proposed zero-shot intent classification method, IntentCapsNet, has been shown to achieve state-of-the-art performance.  ...  Acknowledgments We would like to thank the anonymous reviewers for their helpful comments towards improving the manuscript. This research was supported by the grant HK ITF UIM/377.  ... 
doi:10.18653/v1/d19-1486 dblp:conf/emnlp/LiuZFFLWL19 fatcat:jxfxdsjkmnhjziz3vohqn67cri

El Volumen Louder Por Favor: Code-switching in Task-oriented Semantic Parsing [article]

Arash Einolghozati, Abhinav Arora, Lorena Sainz-Maza Lecanda, Anuj Kumar, Sonal Gupta
2021 arXiv   pre-print
Being able to parse code-switched (CS) utterances, such as Spanish+English or Hindi+English, is essential to democratize task-oriented semantic parsing systems for certain locales.  ...  We propose two data augmentation methods for the zero-shot and the few-shot settings: fine-tune using translate-and-align and augment using a generation model followed by match-and-filter.  ...  In this paper, we first focus on the zero-shot setup for which we only use EN data for the same task domains (we call this in-domain EN data).  ... 
arXiv:2101.10524v3 fatcat:u5gsaj7arjhvpbf7e5o44bk7bq

Zero-Shot Action Recognition from Diverse Object-Scene Compositions [article]

Carlo Bretti, Pascal Mettes
2021 arXiv   pre-print
For this challenging scenario, the current leading approach is to transfer knowledge from the image domain by recognizing objects in videos using pre-trained networks, followed by a semantic matching between  ...  While simple, our composition-based approach outperforms object-based approaches and even state-of-the-art zero-shot approaches that rely on large-scale video datasets with hundreds of seen actions for  ...  Akin to zero-shot learning in the image domain [21, 53] , a wide range of works have shown the ability to recognize unseen actions in videos by learning a shared embedding based on seen training actions  ... 
arXiv:2110.13479v1 fatcat:u4cq4cul3faqdftfjwtfnkjf6i

Towards Zero-Shot Frame Semantic Parsing for Domain Scaling [article]

Ankur Bapna, Gokhan Tur, Dilek Hakkani-Tur, Larry Heck
2017 arXiv   pre-print
State-of-the-art slot filling models for goal-oriented human/machine conversational language understanding systems rely on deep learning methods.  ...  While multi-task training of such models alleviates the need for large in-domain annotated datasets, bootstrapping a semantic parsing model for a new domain using only the semantic frame, such as the back-end  ...  Both LSTM layers are shared across all domains, fol- Zero-Shot Concept Tagging Model The main idea behind the zero-shot concept tagger is to leverage the slot names or descriptions in a domain-agnostic  ... 
arXiv:1707.02363v1 fatcat:bs66sqrzajhrxinxifvx5y6uaq

Attention-Informed Mixed-Language Training for Zero-shot Cross-lingual Task-oriented Dialogue Systems [article]

Zihan Liu, Genta Indra Winata, Zhaojiang Lin, Peng Xu, Pascale Fung
2019 arXiv   pre-print
It leverages very few task-related parallel word pairs to generate code-switching sentences for learning the inter-lingual semantics across languages.  ...  In order to circumvent the expensive and time-consuming data collection, we introduce Attention-Informed Mixed-Language Training (MLT), a novel zero-shot adaptation method for cross-lingual task-oriented  ...  Currently, a few studies have been performed on the zero-shot learning in task-oriented dialogue systems Schuster et al. 2019) .  ... 
arXiv:1911.09273v1 fatcat:bazx4femujbntnwgb4j6bjmfxm

Open cross-domain visual search

William Thong, Pascal Mettes, Cees G.M. Snoek
2020 Computer Vision and Image Understanding  
Open cross-domain visual search is then performed by searching in the common semantic space, regardless of which domains are used as source or target.  ...  We formulate the search as a mapping from every visual domain to a common semantic space, where categories are represented by hyperspherical prototypes.  ...  First, we outline how to represent categories in the semantic embedding space. Second, we propose a mapping function for every domain to the common semantic embedding space.  ... 
doi:10.1016/j.cviu.2020.103045 fatcat:fpwc73iq6bhcre4gzpvuipfvl4

Open Cross-Domain Visual Search [article]

William Thong, Pascal Mettes, Cees G.M. Snoek
2020 arXiv   pre-print
Open cross-domain visual search is then performed by searching in the common semantic space, regardless of which domains are used as source or target.  ...  We formulate the search as a mapping from every visual domain to a common semantic space, where categories are represented by hyperspherical prototypes.  ...  Acknowledgements We thank Herke van Hoof for initial insight, Qing Liu for helpful correspondence, as well as Zenglin Shi and Hubert Banville for feedback.  ... 
arXiv:1911.08621v2 fatcat:gbdy2ltxufa37jop47mjrvtegi

Semantically Tied Paired Cycle Consistency for Zero-Shot Sketch-based Image Retrieval [article]

Anjan Dutta, Zeynep Akata
2019 arXiv   pre-print
In this work, we propose a semantically aligned paired cycle-consistent generative (SEM-PCYC) model for zero-shot SBIR, where each branch maps the visual information to a common semantic space via an adversarial  ...  Existing works either require aligned sketch-image pairs or inefficient memory fusion layer for mapping the visual information to a semantic space.  ...  The Titan Xp and Titan V used for this research were donated by the NVIDIA Corporation.  ... 
arXiv:1903.03372v1 fatcat:ig3bjce4kfeefjnqyhgmujp7dy

Attention-Informed Mixed-Language Training for Zero-Shot Cross-Lingual Task-Oriented Dialogue Systems

Zihan Liu, Genta Indra Winata, Zhaojiang Lin, Peng Xu, Pascale Fung
2020 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
It leverages very few task-related parallel word pairs to generate code-switching sentences for learning the inter-lingual semantics across languages.  ...  In order to circumvent the expensive and time-consuming data collection, we introduce Attention-Informed Mixed-Language Training (MLT), a novel zero-shot adaptation method for cross-lingual task-oriented  ...  Currently, a few studies have been performed on the zero-shot learning in task-oriented dialogue systems Schuster et al. 2019) .  ... 
doi:10.1609/aaai.v34i05.6362 fatcat:h322k7c6zfa5rnfy2g5vnxs2pe
« Previous Showing results 1 — 15 out of 4,168 results