Paper Group NANR 246
Exploring and Enhancing the Transferability of Adversarial Examples. Deep Learning 3D Shapes Using Alt-az Anisotropic 2-Sphere Convolution. DelibGAN: Coarse-to-Fine Text Generation via Adversarial Network. Improving Neural Entity Disambiguation with Graph Embeddings. Semantics and Homothetic Clustering of Hafez Poetry. Selective Convolutional Units …
Exploring and Enhancing the Transferability of Adversarial Examples
| Title | Exploring and Enhancing the Transferability of Adversarial Examples |
| Authors | Lei Wu, Zhanxing Zhu, Cheng Tai |
| Abstract | State-of-the-art deep neural networks are vulnerable to adversarial examples, formed by applying small but malicious perturbations to the original inputs. Moreover, the perturbations can \textit{transfer across models}: adversarial examples generated for a specific model will often mislead other unseen models. Consequently the adversary can leverage it to attack deployed systems without any query, which severely hinders the application of deep learning, especially in the safety-critical areas. In this work, we empirically study how two classes of factors those might influence the transferability of adversarial examples. One is about model-specific factors, including network architecture, model capacity and test accuracy. The other is the local smoothness of loss surface for constructing adversarial examples. Inspired by these understandings on the transferability of adversarial examples, we then propose a simple but effective strategy to enhance the transferability, whose effectiveness is confirmed by a variety of experiments on both CIFAR-10 and ImageNet datasets. |
| Tasks | |
| Published | 2019-05-01 |
| URL | https://openreview.net/forum?id=HyMRUiC9YX |
| https://openreview.net/pdf?id=HyMRUiC9YX | |
| PWC | https://paperswithcode.com/paper/exploring-and-enhancing-the-transferability |
| Repo | |
| Framework | |
Deep Learning 3D Shapes Using Alt-az Anisotropic 2-Sphere Convolution
| Title | Deep Learning 3D Shapes Using Alt-az Anisotropic 2-Sphere Convolution |
| Authors | Min Liu, Fupin Yao, Chiho Choi, Sinha Ayan, Karthik Ramani |
| Abstract | The ground-breaking performance obtained by deep convolutional neural networks (CNNs) for image processing tasks is inspiring research efforts attempting to extend it for 3D geometric tasks. One of the main challenge in applying CNNs to 3D shape analysis is how to define a natural convolution operator on non-euclidean surfaces. In this paper, we present a method for applying deep learning to 3D surfaces using their spherical descriptors and alt-az anisotropic convolution on 2-sphere. A cascade set of geodesic disk filters rotate on the 2-sphere and collect spherical patterns and so to extract geometric features for various 3D shape analysis tasks. We demonstrate theoretically and experimentally that our proposed method has the possibility to bridge the gap between 2D images and 3D shapes with the desired rotation equivariance/invariance, and its effectiveness is evaluated in applications of non-rigid/ rigid shape classification and shape retrieval. |
| Tasks | 3D Shape Analysis |
| Published | 2019-05-01 |
| URL | https://openreview.net/forum?id=rkeSiiA5Fm |
| https://openreview.net/pdf?id=rkeSiiA5Fm | |
| PWC | https://paperswithcode.com/paper/deep-learning-3d-shapes-using-alt-az |
| Repo | |
| Framework | |
DelibGAN: Coarse-to-Fine Text Generation via Adversarial Network
| Title | DelibGAN: Coarse-to-Fine Text Generation via Adversarial Network |
| Authors | Ke Wang, Xiaojun Wan |
| Abstract | In this paper, we propose a novel adversarial learning framework, namely DelibGAN, for generating high-quality sentences without supervision. Our framework consists of a coarse-to-fine generator, which contains a first-pass decoder and a second-pass decoder, and a multiple instance discriminator. And we propose two training mechanisms DelibGAN-I and DelibGAN-II. The discriminator is used to fine-tune the second-pass decoder in DelibGAN-I and further evaluate the importance of each word and tune the first-pass decoder in DelibGAN-II. We compare our models with several typical and state-of-the-art unsupervised generic text generation models on three datasets (a synthetic dataset, a descriptive text dataset and a sentimental text dataset). Both qualitative and quantitative experimental results show that our models produce more realistic samples, and DelibGAN-II performs best. |
| Tasks | Text Generation |
| Published | 2019-05-01 |
| URL | https://openreview.net/forum?id=SkMx_iC9K7 |
| https://openreview.net/pdf?id=SkMx_iC9K7 | |
| PWC | https://paperswithcode.com/paper/delibgan-coarse-to-fine-text-generation-via |
| Repo | |
| Framework | |
Improving Neural Entity Disambiguation with Graph Embeddings
| Title | Improving Neural Entity Disambiguation with Graph Embeddings |
| Authors | {"O}zge Sevgili, Alex Panchenko, er, Chris Biemann |
| Abstract | Entity Disambiguation (ED) is the task of linking an ambiguous entity mention to a corresponding entry in a knowledge base. Current methods have mostly focused on unstructured text data to learn representations of entities, however, there is structured information in the knowledge base itself that should be useful to disambiguate entities. In this work, we propose a method that uses graph embeddings for integrating structured information from the knowledge base with unstructured information from text-based representations. Our experiments confirm that graph embeddings trained on a graph of hyperlinks between Wikipedia articles improve the performances of simple feed-forward neural ED model and a state-of-the-art neural ED system. |
| Tasks | Entity Disambiguation |
| Published | 2019-07-01 |
| URL | https://www.aclweb.org/anthology/P19-2044/ |
| https://www.aclweb.org/anthology/P19-2044 | |
| PWC | https://paperswithcode.com/paper/improving-neural-entity-disambiguation-with |
| Repo | |
| Framework | |
Semantics and Homothetic Clustering of Hafez Poetry
| Title | Semantics and Homothetic Clustering of Hafez Poetry |
| Authors | Arya Rahgozar, Diana Inkpen |
| Abstract | We have created two sets of labels for Hafez (1315-1390) poems, using unsupervised learning. Our labels are the only semantic clustering alternative to the previously existing, hand-labeled, gold-standard classification of Hafez poems, to be used for literary research. We have cross-referenced, measured and analyzed the agreements of our clustering labels with Houman{'}s chronological classes. Our features are based on topic modeling and word embeddings. We also introduced a similarity of similarities{'} features, we called homothetic clustering approach that proved effective, in case of Hafez{'}s small corpus of ghazals2. Although all our experiments showed different clusters when compared with Houman{'}s classes, we think they were valid in their own right to have provided further insights, and have proved useful as a contrasting alternative to Houman{'}s classes. Our homothetic clusterer and its feature design and engineering framework can be used for further semantic analysis of Hafez{'}s poetry and other similar literary research. |
| Tasks | Word Embeddings |
| Published | 2019-06-01 |
| URL | https://www.aclweb.org/anthology/W19-2511/ |
| https://www.aclweb.org/anthology/W19-2511 | |
| PWC | https://paperswithcode.com/paper/semantics-and-homothetic-clustering-of-hafez |
| Repo | |
| Framework | |
Selective Convolutional Units: Improving CNNs via Channel Selectivity
| Title | Selective Convolutional Units: Improving CNNs via Channel Selectivity |
| Authors | Jongheon Jeong, Jinwoo Shin |
| Abstract | Bottleneck structures with identity (e.g., residual) connection are now emerging popular paradigms for designing deep convolutional neural networks (CNN), for processing large-scale features efficiently. In this paper, we focus on the information-preserving nature of identity connection and utilize this to enable a convolutional layer to have a new functionality of channel-selectivity, i.e., re-distributing its computations to important channels. In particular, we propose Selective Convolutional Unit (SCU), a widely-applicable architectural unit that improves parameter efficiency of various modern CNNs with bottlenecks. During training, SCU gradually learns the channel-selectivity on-the-fly via the alternative usage of (a) pruning unimportant channels, and (b) rewiring the pruned parameters to important channels. The rewired parameters emphasize the target channel in a way that selectively enlarges the convolutional kernels corresponding to it. Our experimental results demonstrate that the SCU-based models without any postprocessing generally achieve both model compression and accuracy improvement compared to the baselines, consistently for all tested architectures. |
| Tasks | Model Compression |
| Published | 2019-05-01 |
| URL | https://openreview.net/forum?id=SJlt6oA9Fm |
| https://openreview.net/pdf?id=SJlt6oA9Fm | |
| PWC | https://paperswithcode.com/paper/selective-convolutional-units-improving-cnns |
| Repo | |
| Framework | |
Exploring Adequacy Errors in Neural Machine Translation with the Help of Cross-Language Aligned Word Embeddings
| Title | Exploring Adequacy Errors in Neural Machine Translation with the Help of Cross-Language Aligned Word Embeddings |
| Authors | Michael Ustaszewski |
| Abstract | Neural machine translation (NMT) was shown to produce more fluent output than phrase-based statistical (PBMT) and rule-based machine translation (RBMT). However, improved fluency makes it more difficult for post editors to identify and correct adequacy errors, because unlike RBMT and SMT, in NMT adequacy errors are frequently not anticipated by fluency errors. Omissions and additions of content in otherwise flawlessly fluent NMT output are the most prominent types of such adequacy errors, which can only be detected with reference to source texts. This contribution explores the degree of semantic similarity between source texts, NMT output and post edited output. In this way, computational semantic similarity scores (cosine similarity) are related to human quality judgments. The analyses are based on publicly available NMT post editing data annotated for errors in three language pairs (EN-DE, EN-LV, EN-HR) with the Multidimensional Quality Metrics (MQM). Methodologically, this contribution tests whether cross-language aligned word embeddings as the sole source of semantic information mirror human error annotation. |
| Tasks | Machine Translation, Semantic Similarity, Semantic Textual Similarity, Word Embeddings |
| Published | 2019-09-01 |
| URL | https://www.aclweb.org/anthology/W19-8715/ |
| https://www.aclweb.org/anthology/W19-8715 | |
| PWC | https://paperswithcode.com/paper/exploring-adequacy-errors-in-neural-machine |
| Repo | |
| Framework | |
JCTICOL at SemEval-2019 Task 6: Classifying Offensive Language in Social Media using Deep Learning Methods, Word/Character N-gram Features, and Preprocessing Methods
| Title | JCTICOL at SemEval-2019 Task 6: Classifying Offensive Language in Social Media using Deep Learning Methods, Word/Character N-gram Features, and Preprocessing Methods |
| Authors | Yaakov HaCohen-Kerner, Ziv Ben-David, Gal Didi, Eli Cahn, Shalom Rochman, Elyashiv Shayovitz |
| Abstract | In this paper, we describe our submissions to SemEval-2019 task 6 contest. We tackled all three sub-tasks in this task {``}OffensEval - Identifying and Categorizing Offensive Language in Social Media{''}. In our system called JCTICOL (Jerusalem College of Technology Identifies and Categorizes Offensive Language), we applied various supervised ML methods. We applied various combinations of word/character n-gram features using the TF-IDF scheme. In addition, we applied various combinations of seven basic preprocessing methods. Our best submission, an RNN model was ranked at the 25th position out of 65 submissions for the most complex sub-task (C). | |
| Tasks | |
| Published | 2019-06-01 |
| URL | https://www.aclweb.org/anthology/S19-2115/ |
| https://www.aclweb.org/anthology/S19-2115 | |
| PWC | https://paperswithcode.com/paper/jcticol-at-semeval-2019-task-6-classifying |
| Repo | |
| Framework | |
Physiological Signal Embeddings (PHASE) via Interpretable Stacked Models
| Title | Physiological Signal Embeddings (PHASE) via Interpretable Stacked Models |
| Authors | Hugh Chen, Scott Lundberg, Gabe Erion, Su-In Lee |
| Abstract | In health, machine learning is increasingly common, yet neural network embedding (representation) learning is arguably under-utilized for physiological signals. This inadequacy stands out in stark contrast to more traditional computer science domains, such as computer vision (CV), and natural language processing (NLP). For physiological signals, learning feature embeddings is a natural solution to data insufficiency caused by patient privacy concerns – rather than share data, researchers may share informative embedding models (i.e., representation models), which map patient data to an output embedding. Here, we present the PHASE (PHysiologicAl Signal Embeddings) framework, which consists of three components: i) learning neural network embeddings of physiological signals, ii) predicting outcomes based on the learned embedding, and iii) interpreting the prediction results by estimating feature attributions in the “stacked” models (i.e., feature embedding model followed by prediction model). PHASE is novel in three ways: 1) To our knowledge, PHASE is the first instance of transferal of neural networks to create physiological signal embeddings. 2) We present a tractable method to obtain feature attributions through stacked models. We prove that our stacked model attributions can approximate Shapley values – attributions known to have desirable properties – for arbitrary sets of models. 3) PHASE was extensively tested in a cross-hospital setting including publicly available data. In our experiments, we show that PHASE significantly outperforms alternative embeddings – such as raw, exponential moving average/variance, and autoencoder – currently in use. Furthermore, we provide evidence that transferring neural network embedding/representation learners between distinct hospitals still yields performant embeddings and offer recommendations when transference is ineffective. |
| Tasks | Network Embedding, Representation Learning |
| Published | 2019-05-01 |
| URL | https://openreview.net/forum?id=SygInj05Fm |
| https://openreview.net/pdf?id=SygInj05Fm | |
| PWC | https://paperswithcode.com/paper/physiological-signal-embeddings-phase-via |
| Repo | |
| Framework | |
Leveraging BERT to Improve the FEARS Index for Stock Forecasting
| Title | Leveraging BERT to Improve the FEARS Index for Stock Forecasting |
| Authors | Linyi Yang, Ruihai Dong, Tin Lok James Ng, Yang Xu |
| Abstract | |
| Tasks | |
| Published | 2019-08-01 |
| URL | https://www.aclweb.org/anthology/W19-5509/ |
| https://www.aclweb.org/anthology/W19-5509 | |
| PWC | https://paperswithcode.com/paper/leveraging-bert-to-improve-the-fears-index |
| Repo | |
| Framework | |
Learning Unsupervised Learning Rules
| Title | Learning Unsupervised Learning Rules |
| Authors | Luke Metz, Niru Maheswaranathan, Brian Cheung, Jascha Sohl-Dickstein |
| Abstract | A major goal of unsupervised learning is to discover data representations that are useful for subsequent tasks, without access to supervised labels during training. Typically, this goal is approached by minimizing a surrogate objective, such as the negative log likelihood of a generative model, with the hope that representations useful for subsequent tasks will arise incidentally. In this work, we propose instead to directly target a later desired task by meta-learning an unsupervised learning rule, which leads to representations useful for that task. Here, our desired task (meta-objective) is the performance of the representation on semi-supervised classification, and we meta-learn an algorithm – an unsupervised weight update rule – that produces representations that perform well under this meta-objective. Additionally, we constrain our unsupervised update rule to a be a biologically-motivated, neuron-local function, which enables it to generalize to novel neural network architectures. We show that the meta-learned update rule produces useful features and sometimes outperforms existing unsupervised learning techniques. We further show that the meta-learned unsupervised update rule generalizes to train networks with different widths, depths, and nonlinearities. It also generalizes to train on data with randomly permuted input dimensions and even generalizes from image datasets to a text task. |
| Tasks | Meta-Learning |
| Published | 2019-05-01 |
| URL | https://openreview.net/forum?id=HkNDsiC9KQ |
| https://openreview.net/pdf?id=HkNDsiC9KQ | |
| PWC | https://paperswithcode.com/paper/learning-unsupervised-learning-rules |
| Repo | |
| Framework | |
ShieldNets: Defending Against Adversarial Attacks Using Probabilistic Adversarial Robustness
| Title | ShieldNets: Defending Against Adversarial Attacks Using Probabilistic Adversarial Robustness |
| Authors | Rajkumar Theagarajan, Ming Chen, Bir Bhanu, Jing Zhang |
| Abstract | Defending adversarial attack is a critical step towards reliable deployment of deep learning empowered solutions for industrial applications. Probabilistic adversarial robustness (PAR), as a theoretical framework, is introduced to neutralize adversarial attacks by concentrating sample probability to adversarial-free zones. Distinct to most of the existing defense mechanisms that require modifying the architecture/training of the target classifier which is not feasible in the real-world scenario, e.g., when a model has already been deployed, PAR is designed in the first place to provide proactive protection to an existing fixed model. ShieldNet is implemented as a demonstration of PAR in this work by using PixelCNN. Experimental results show that this approach is generalizable, robust against adversarial transferability and resistant to a wide variety of attacks on the Fashion-MNIST and CIFAR10 datasets, respectively. |
| Tasks | Adversarial Attack |
| Published | 2019-06-01 |
| URL | http://openaccess.thecvf.com/content_CVPR_2019/html/Theagarajan_ShieldNets_Defending_Against_Adversarial_Attacks_Using_Probabilistic_Adversarial_Robustness_CVPR_2019_paper.html |
| http://openaccess.thecvf.com/content_CVPR_2019/papers/Theagarajan_ShieldNets_Defending_Against_Adversarial_Attacks_Using_Probabilistic_Adversarial_Robustness_CVPR_2019_paper.pdf | |
| PWC | https://paperswithcode.com/paper/shieldnets-defending-against-adversarial |
| Repo | |
| Framework | |
Improving Knowledge Base Construction from Robust Infobox Extraction
| Title | Improving Knowledge Base Construction from Robust Infobox Extraction |
| Authors | Boya Peng, Yejin Huh, Xiao Ling, Michele Banko |
| Abstract | A capable, automatic Question Answering (QA) system can provide more complete and accurate answers using a comprehensive knowledge base (KB). One important approach to constructing a comprehensive knowledge base is to extract information from Wikipedia infobox tables to populate an existing KB. Despite previous successes in the Infobox Extraction (IBE) problem (e.g., DBpedia), three major challenges remain: 1) Deterministic extraction patterns used in DBpedia are vulnerable to template changes; 2) Over-trusting Wikipedia anchor links can lead to entity disambiguation errors; 3) Heuristic-based extraction of unlinkable entities yields low precision, hurting both accuracy and completeness of the final KB. This paper presents a robust approach that tackles all three challenges. We build probabilistic models to predict relations between entity mentions directly from the infobox tables in HTML. The entity mentions are linked to identifiers in an existing KB if possible. The unlinkable ones are also parsed and preserved in the final output. Training data for both the relation extraction and the entity linking models are automatically generated using distant supervision. We demonstrate the empirical effectiveness of the proposed method in both precision and recall compared to a strong IBE baseline, DBpedia, with an absolute improvement of 41.3{%} in average F1. We also show that our extraction makes the final KB significantly more complete, improving the completeness score of list-value relation types by 61.4{%}. |
| Tasks | Entity Disambiguation, Entity Linking, Question Answering, Relation Extraction |
| Published | 2019-06-01 |
| URL | https://www.aclweb.org/anthology/N19-2018/ |
| https://www.aclweb.org/anthology/N19-2018 | |
| PWC | https://paperswithcode.com/paper/improving-knowledge-base-construction-from |
| Repo | |
| Framework | |
TDBot at SemEval-2019 Task 3: Context Aware Emotion Detection Using A Conditioned Classification Approach
| Title | TDBot at SemEval-2019 Task 3: Context Aware Emotion Detection Using A Conditioned Classification Approach |
| Authors | Sourabh Maity |
| Abstract | With the system description it is shown how to use the context information while detecting the emotion in a dialogue. Some guidelines about how to handle emojis was also laid out. While developing this system I realized the importance of pre-processing in conversational text data, or in general NLP related tasks; it can not be over emphasized. |
| Tasks | |
| Published | 2019-06-01 |
| URL | https://www.aclweb.org/anthology/S19-2058/ |
| https://www.aclweb.org/anthology/S19-2058 | |
| PWC | https://paperswithcode.com/paper/tdbot-at-semeval-2019-task-3-context-aware |
| Repo | |
| Framework | |
Proceedings of the 3rd Workshop on Evaluating Vector Space Representations for NLP
| Title | Proceedings of the 3rd Workshop on Evaluating Vector Space Representations for NLP |
| Authors | |
| Abstract | |
| Tasks | |
| Published | 2019-06-01 |
| URL | https://www.aclweb.org/anthology/W19-2000/ |
| https://www.aclweb.org/anthology/W19-2000 | |
| PWC | https://paperswithcode.com/paper/proceedings-of-the-3rd-workshop-on-evaluating |
| Repo | |
| Framework | |
