Filters








235 Hits in 6.4 sec

Supervised Visual Attention for Multimodal Neural Machine Translation

Tetsuro Nishihara, Akihiro Tamura, Takashi Ninomiya, Yutaro Omote, Hideki Nakayama
2021 Journal of Natural Language Processing  
Our experiments on English-German and German-English translation tasks using the Multi30k dataset and on English-Japanese and Japanese-English translation tasks using the Flickr30k Entities JP dataset  ...  This paper proposed a supervised visual attention mechanism for multimodal neural machine translation (MNMT), trained with constraints based on manual alignments between words in a sentence and their corresponding  ...  In 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings.  ... 
doi:10.5715/jnlp.28.554 fatcat:gp6zorbdeje5jn2gi6sqlfifam

PARL: Enhancing Diversity of Ensemble Networks to Resist Adversarial Attacks via Pairwise Adversarially Robust Loss Function [article]

Manaar Alam, Shubhajit Datta, Debdeep Mukhopadhyay, Arijit Mondal, Partha Pratim Chakrabarti
2021 arXiv   pre-print
The security of Deep Learning classifiers is a critical field of study because of the existence of adversarial attacks.  ...  Such attacks usually rely on the principle of transferability, where an adversarial example crafted on a surrogate classifier tends to mislead the target classifier trained on the same dataset even if  ...  In 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings, 2015.  ... 
arXiv:2112.04948v1 fatcat:i7ab4hvgprcgvdpnowrlwmaiwa

Explainable Abstract Trains Dataset [article]

Manuel de Sousa Ribeiro, Ludwig Krippahl, Joao Leite
2020 arXiv   pre-print
The Explainable Abstract Trains Dataset is an image dataset containing simplified representations of trains.  ...  The dataset is accompanied by an ontology that conceptualizes and classifies the depicted trains based on their visual characteristics, allowing for a precise understanding of how each train was labeled  ...  In Yoshua Bengio and Yann LeCun, editors, 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings, 2015.  ... 
arXiv:2012.12115v1 fatcat:deixf5bzajf2rnktjk36q2rwhm

Interpretable Word Embeddings via Informative Priors [article]

Miriam Hurtado Bodell, Martin Arvidsson, Måns Magnusson
2019 arXiv   pre-print
Word embeddings have demonstrated strong performance on NLP tasks.  ...  Experimental results show that sensible priors can capture latent semantic concepts better than or on-par with the current state of the art, while retaining the simplicity and generalizability of using  ...  In 3rd Inter- national Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings.  ... 
arXiv:1909.01459v1 fatcat:53ancapz4bdbrjf3d7qxf5ewzu

Lexicon-constrained Copying Network for Chinese Abstractive Summarization [article]

Boyan Wan, Mishal Sohail
2021 arXiv   pre-print
On the source side, words and characters are aggregated into the same input memory using a Transformerbased encoder.  ...  Experiments results on a Chinese social media dataset show that our model can work standalone or with the word selector.  ...  Bengio, “Neural machine translation by jointly learning to align and translate,” in 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015  ... 
arXiv:2010.08197v2 fatcat:7mp7mvrh2jbqzpch4zp7axfoce

Fast Test Input Generation for Finding Deviated Behaviors in Compressed Deep Neural Network [article]

Yongqiang Tian, Wuqi Zhang, Ming Wen, Shing-Chi Cheung, Chengnian Sun, Shiqing Ma, Yu Jiang
2021 arXiv   pre-print
We evaluated TriggerFinder on 18 compressed models with two datasets.  ...  However, compressed models usually hide their architecture and gradient information; without such internal information as guidance, it becomes difficult to effectively and efficiently trigger deviated  ...  TriggerFinder incorpo- ing Representations, ICLR 2015, San Diego, CA, USA, May rates a novel fitness function to determine whether to use 7-9, 2015, Conference Track Proceedings. a mutated input  ... 
arXiv:2112.02819v1 fatcat:qgfj2sfazfc2dd4b2aqhydqpgu

Energy-based Self-attentive Learning of Abstractive Communities for Spoken Language Understanding [article]

Guokan Shang, Michalis Vazirgiannis École Polytechnique,
2019 arXiv   pre-print
Experiments on the AMI corpus show that our system outperforms multiple energy-based and non-energy based baselines from the state-of-the-art. Code and data are publicly available.  ...  In 3rd Inter- national Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings.  ...  In 3rd International Confer- ence on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Workshop Track Proceedings. Diederik P. Kingma and Jimmy Ba. 2015.  ... 
arXiv:1904.09491v2 fatcat:yepg4w45rzdhtdkrf2evlt343y

Few-shot and Zero-shot Approaches to Legal Text Classification: A Case Study in the Financial Sector

Rajdeep Sarkar, Atul Kr. Ojha, Jay Megaro, John Mariano, Vall Herard, John P. McCrae
2021 Zenodo  
, San Diego, CA, USA, May 7-9, 2015, 2017, volume 2143 of CEUR Workshop Proceedings.  ...  In 3rd Inter- 16th International Conference on Artificial Intelli- national Conference on Learning Representations, gence and Law (ICAIL 2017), London, UK, June 16, ICLR 2015  ... 
doi:10.5281/zenodo.5772042 fatcat:sgkfcnqtmjainnvs3fyoqi4z2m

A Transformer-based Neural Model for Chinese Word Segmentation and Part-of-Speech Tagging

Xinxin Li
2021 IJARCCE  
Bengio, “Neural machine translation by jointly learning to align and translate,” in 3rd international conference on learning representations, ICLR 2015, san diego, CA, USA, may 7-9, 2015, conference  ...  track proceedings, 2015. [18].  ... 
doi:10.17148/ijarcce.2021.101201 fatcat:hctjxq7flrcs3fessowd3hiehy

Which images to label for few-shot medical landmark detection? [article]

Quan Quan, Qingsong Yao, Jun Li, S. Kevin Zhou
2021 arXiv   pre-print
The success of deep learning methods relies on the availability of well-labeled large-scale datasets.  ...  However, a crucial yet previously overlooked problem in few-shot learning is about the selection of template images for annotation before learning, which affects the final performance.  ...  IEEE Computer Society, 2017. 5 Diego, CA, USA, May 7-9, 2015, Conference Track Proceed- [41] Qingsong Yao, Zecheng He, Hu Han, and S Kevin Zhou.  ... 
arXiv:2112.04386v2 fatcat:ajn62lhsz5bl7hruzmujcntc3u

Combining Textual Features for the Detection of Hateful and Offensive Language [article]

Sherzod Hakimov, Ralph Ewerth
2021 arXiv   pre-print
In this paper, we present an analysis of combining different textual features for the detection of hateful or offensive posts on Twitter.  ...  The detection of offensive, hateful and profane language has become a critical challenge since many users in social networks are exposed to cyberbullying activities on a daily basis.  ...  LeCun (Eds.), 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings, 2015.  ... 
arXiv:2112.04803v1 fatcat:nwvs4vongnhubk2thtk2xlwysu

Semantic Representation for Dialogue Modeling [article]

Xuefeng Bai, Yulong Chen, Linfeng Song, Yue Zhang
2021 arXiv   pre-print
Experimental results on both dialogue understanding and response generation tasks show the superiority of our model.  ...  To our knowledge, we are the first to leverage a formal semantic representation into neural dialogue modeling.  ...  Dialogue relation ex- ICLR 2015, San Diego, CA, USA, May 7-9, 2015, traction with document-level heterogeneous graph Conference Track Proceedings.  ... 
arXiv:2105.10188v2 fatcat:unzjukziljbc7lqs4w722gczh4

Code Summarization with Structure-induced Transformer [article]

Hongqiu Wu and Hai Zhao and Min Zhang
2021 arXiv   pre-print
Extensive experiments show that our proposed structure-induced Transformer helps achieve new state-of-the-art results on benchmarks.  ...  Thus previous works attempt to apply structure-based traversal (SBT) or non-sequential models like Tree-LSTM and graph neural network (GNN) to learn structural program semantics.  ...  In 3rd Inter- national Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings. Antonio Valerio Miceli Barone and Rico Sennrich. 2017.  ... 
arXiv:2012.14710v2 fatcat:qshoiyhanbbpdp4k7ktgaopmv4

Variational Neural Machine Translation with Normalizing Flows [article]

Hendra Setiawan, Matthias Sperber, Udhay Nallasamy, Matthias Paulik
2020 arXiv   pre-print
The latent variable modeling may introduce useful statistical dependencies that can improve translation accuracy.  ...  Unfortunately, learning informative latent variables is non-trivial, as the latent space can be prohibitively large, and the latent codes are prone to be ignored by many translation models at training  ...  In International Conference on Representation Learning (ICLR), San Diego, USA. Hareesh Bahuleyan, Lili Mou, Olga Vech- tomova, and Pascal Poupart. 2018.  ... 
arXiv:2005.13978v1 fatcat:7ysqtt4kgzgm3l4kddct7byihy

Translating Human Mobility Forecasting through Natural Language Generation [article]

Hao Xue, Flora D. Salim, Yongli Ren, Charles L. A. Clarke
2021 arXiv   pre-print
Specifically, it consists of one main branch for language generation and one auxiliary branch to directly learn mobility patterns.  ...  Extensive experiments on three real-world datasets demonstrate that the proposed SHIFT is effective and presents a new revolutionary approach to forecasting human mobility.  ...  In Proceedings of the Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May Web Conference 2021. 2177–2185. 7-9, 2015, Conference Track Proceedings, Yoshua  ... 
arXiv:2112.11481v1 fatcat:ydqmcvjcfjhvnjhrbq3hteeska
« Previous Showing results 1 — 15 out of 235 results