Paper Group NANR 211
Termbase Exchange (TBX). MACH: Embarrassingly parallel $K$-class classification in $O(d\log{K})$ memory and $O(K\log{K} + d\log{K})$ time, instead of $O(Kd)$. Unrestricted Bridging Resolution. An Integrated Representation of Linguistic and Social Functions of Code-Switching. Multi-Fidelity Black-Box Optimization with Hierarchical Partitions. Relati …
Termbase Exchange (TBX)
| Title | Termbase Exchange (TBX) |
| Authors | Sue Wright |
| Abstract | |
| Tasks | Machine Translation |
| Published | 2018-03-01 |
| URL | https://www.aclweb.org/anthology/W18-2002/ |
| https://www.aclweb.org/anthology/W18-2002 | |
| PWC | https://paperswithcode.com/paper/termbase-exchange-tbx |
| Repo | |
| Framework | |
MACH: Embarrassingly parallel $K$-class classification in $O(d\log{K})$ memory and $O(K\log{K} + d\log{K})$ time, instead of $O(Kd)$
| Title | MACH: Embarrassingly parallel $K$-class classification in $O(d\log{K})$ memory and $O(K\log{K} + d\log{K})$ time, instead of $O(Kd)$ |
| Authors | Qixuan Huang, Anshumali Shrivastava, Yiqiu Wang |
| Abstract | We present Merged-Averaged Classifiers via Hashing (MACH) for $K$-classification with large $K$. Compared to traditional one-vs-all classifiers that require $O(Kd)$ memory and inference cost, MACH only need $O(d\log{K})$ memory while only requiring $O(K\log{K} + d\log{K})$ operation for inference. MACH is the first generic $K$-classification algorithm, with provably theoretical guarantees, which requires $O(\log{K})$ memory without any assumption on the relationship between classes. MACH uses universal hashing to reduce classification with a large number of classes to few independent classification task with very small (constant) number of classes. We provide theoretical quantification of accuracy-memory tradeoff by showing the first connection between extreme classification and heavy hitters. With MACH we can train ODP dataset with 100,000 classes and 400,000 features on a single Titan X GPU (12GB), with the classification accuracy of 19.28%, which is the best-reported accuracy on this dataset. Before this work, the best performing baseline is a one-vs-all classifier that requires 40 billion parameters (320 GB model size) and achieves 9% accuracy. In contrast, MACH can achieve 9% accuracy with 480x reduction in the model size (of mere 0.6GB). With MACH, we also demonstrate complete training of fine-grained imagenet dataset (compressed size 104GB), with 21,000 classes, on a single GPU. |
| Tasks | |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=r1RQdCg0W |
| https://openreview.net/pdf?id=r1RQdCg0W | |
| PWC | https://paperswithcode.com/paper/mach-embarrassingly-parallel-k-class |
| Repo | |
| Framework | |
Unrestricted Bridging Resolution
| Title | Unrestricted Bridging Resolution |
| Authors | Yufang Hou, Katja Markert, Michael Strube |
| Abstract | In contrast to identity anaphors, which indicate coreference between a noun phrase and its antecedent, bridging anaphors link to their antecedent(s) via lexico-semantic, frame, or encyclopedic relations. Bridging resolution involves recognizing bridging anaphors and finding links to antecedents. In contrast to most prior work, we tackle both problems. Our work also follows a more wide-ranging definition of bridging than most previous work and does not impose any restrictions on the type of bridging anaphora or relations between anaphor and antecedent. We create a corpus (ISNotes) annotated for information status (IS), bridging being one of the IS subcategories. The annotations reach high reliability for all categories and marginal reliability for the bridging subcategory. We use a two-stage statistical global inference method for bridging resolution. Given all mentions in a document, the first stage, bridging anaphora recognition, recognizes bridging anaphors as a subtask of learning fine-grained IS. We use a cascading collective classification method where (i) collective classification allows us to investigate relations among several mentions and autocorrelation among IS classes and (ii) cascaded classification allows us to tackle class imbalance, important for minority classes such as bridging. We show that our method outperforms current methods both for IS recognition overall as well as for bridging, specifically. The second stage, bridging antecedent selection, finds the antecedents for all predicted bridging anaphors. We investigate the phenomenon of semantically or syntactically related bridging anaphors that share the same antecedent, a phenomenon we call sibling anaphors. We show that taking sibling anaphors into account in a joint inference model improves antecedent selection performance. In addition, we develop semantic and salience features for antecedent selection and suggest a novel method to build the candidate antecedent list for an anaphor, using the discourse scope of the anaphor. Our model outperforms previous work significantly. |
| Tasks | |
| Published | 2018-06-01 |
| URL | https://www.aclweb.org/anthology/J18-2002/ |
| https://www.aclweb.org/anthology/J18-2002 | |
| PWC | https://paperswithcode.com/paper/unrestricted-bridging-resolution |
| Repo | |
| Framework | |
An Integrated Representation of Linguistic and Social Functions of Code-Switching
| Title | An Integrated Representation of Linguistic and Social Functions of Code-Switching |
| Authors | Silvana Hartmann, Monojit Choudhury, Kalika Bali |
| Abstract | |
| Tasks | |
| Published | 2018-05-01 |
| URL | https://www.aclweb.org/anthology/L18-1256/ |
| https://www.aclweb.org/anthology/L18-1256 | |
| PWC | https://paperswithcode.com/paper/an-integrated-representation-of-linguistic |
| Repo | |
| Framework | |
Multi-Fidelity Black-Box Optimization with Hierarchical Partitions
| Title | Multi-Fidelity Black-Box Optimization with Hierarchical Partitions |
| Authors | Rajat Sen, Kirthevasan Kandasamy, Sanjay Shakkottai |
| Abstract | Motivated by settings such as hyper-parameter tuning and physical simulations, we consider the problem of black-box optimization of a function. Multi-fidelity techniques have become popular for applications where exact function evaluations are expensive, but coarse (biased) approximations are available at much lower cost. A canonical example is that of hyper-parameter selection in a learning algorithm. The learning algorithm can be trained for fewer iterations – this results in a lower cost, but its validation error is only coarsely indicative of the same if the algorithm had been trained till completion. We incorporate the multi-fidelity setup into the powerful framework of black-box optimization through hierarchical partitioning. We develop tree-search based multi-fidelity algorithms with theoretical guarantees on simple regret. We finally demonstrate the performance gains of our algorithms on both real and synthetic datasets. |
| Tasks | Physical Simulations |
| Published | 2018-07-01 |
| URL | https://icml.cc/Conferences/2018/Schedule?showEvent=2264 |
| http://proceedings.mlr.press/v80/sen18a/sen18a.pdf | |
| PWC | https://paperswithcode.com/paper/multi-fidelity-black-box-optimization-with |
| Repo | |
| Framework | |
Relational Multi-Instance Learning for Concept Annotation from Medical Time Series
| Title | Relational Multi-Instance Learning for Concept Annotation from Medical Time Series |
| Authors | Sanjay Purushotham, Zhengping Che, Bo Jiang, Tanachat Nilanon, Yan Liu |
| Abstract | Recent advances in computing technology and sensor design have made it easier to collect longitudinal or time series data from patients, resulting in a gigantic amount of available medical data. Most of the medical time series lack annotations or even when the annotations are available they could be subjective and prone to human errors. Earlier works have developed natural language processing techniques to extract concept annotations and/or clinical narratives from doctor notes. However, these approaches are slow and do not use the accompanying medical time series data. To address this issue, we introduce the problem of concept annotation for the medical time series data, i.e., the task of predicting and localizing medical concepts by using the time series data as input. We propose Relational Multi-Instance Learning (RMIL) - a deep Multi Instance Learning framework based on recurrent neural networks, which uses pooling functions and attention mechanisms for the concept annotation tasks. Empirical results on medical datasets show that our proposed models outperform various multi-instance learning models. |
| Tasks | Time Series |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=ByJbJwxCW |
| https://openreview.net/pdf?id=ByJbJwxCW | |
| PWC | https://paperswithcode.com/paper/relational-multi-instance-learning-for |
| Repo | |
| Framework | |
Building a List of Synonymous Words and Phrases of Japanese Compound Verbs
| Title | Building a List of Synonymous Words and Phrases of Japanese Compound Verbs |
| Authors | Kyoko Kanzaki, Hitoshi Isahara |
| Abstract | |
| Tasks | Morphological Analysis, Word Sense Disambiguation |
| Published | 2018-05-01 |
| URL | https://www.aclweb.org/anthology/L18-1376/ |
| https://www.aclweb.org/anthology/L18-1376 | |
| PWC | https://paperswithcode.com/paper/building-a-list-of-synonymous-words-and |
| Repo | |
| Framework | |
WordNet-Shp: Towards the Building of a Lexical Database for a Peruvian Minority Language
| Title | WordNet-Shp: Towards the Building of a Lexical Database for a Peruvian Minority Language |
| Authors | Diego Magui{~n}o-Valencia, Arturo Oncevay-Marcos, Marco A. Sobrevilla Cabezudo |
| Abstract | |
| Tasks | Machine Translation, Word Sense Disambiguation |
| Published | 2018-05-01 |
| URL | https://www.aclweb.org/anthology/L18-1697/ |
| https://www.aclweb.org/anthology/L18-1697 | |
| PWC | https://paperswithcode.com/paper/wordnet-shp-towards-the-building-of-a-lexical |
| Repo | |
| Framework | |
Thinking like a machine — generating visual rationales through latent space optimization
| Title | Thinking like a machine — generating visual rationales through latent space optimization |
| Authors | Jarrel Seah, Jennifer Tang, Andy Kitchen, Jonathan Seah |
| Abstract | Interpretability and small labelled datasets are key issues in the practical application of deep learning, particularly in areas such as medicine. In this paper, we present a semi-supervised technique that addresses both these issues simultaneously. We learn dense representations from large unlabelled image datasets, then use those representations to both learn classifiers from small labeled sets and generate visual rationales explaining the predictions. Using chest radiography diagnosis as a motivating application, we show our method has good generalization ability by learning to represent our chest radiography dataset while training a classifier on an separate set from a different institution. Our method identifies heart failure and other thoracic diseases. For each prediction, we generate visual rationales for positive classifications by optimizing a latent representation to minimize the probability of disease while constrained by a similarity measure in image space. Decoding the resultant latent representation produces an image without apparent disease. The difference between the original and the altered image forms an interpretable visual rationale for the algorithm’s prediction. Our method simultaneously produces visual rationales that compare favourably to previous techniques and a classifier that outperforms the current state-of-the-art. |
| Tasks | |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=B13EC5u6W |
| https://openreview.net/pdf?id=B13EC5u6W | |
| PWC | https://paperswithcode.com/paper/thinking-like-a-machine-generating-visual |
| Repo | |
| Framework | |
How do deep convolutional neural networks learn from raw audio waveforms?
| Title | How do deep convolutional neural networks learn from raw audio waveforms? |
| Authors | Yuan Gong, Christian Poellabauer |
| Abstract | Prior work on speech and audio processing has demonstrated the ability to obtain excellent performance when learning directly from raw audio waveforms using convolutional neural networks (CNNs). However, the exact inner workings of a CNN remain unclear, which hinders further developments and improvements into this direction. In this paper, we theoretically analyze and explain how deep CNNs learn from raw audio waveforms and identify potential limitations of existing network structures. Based on this analysis, we further propose a new network architecture (called SimpleNet), which offers a very simple but concise structure and high model interpretability. |
| Tasks | |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=S1Ow_e-Rb |
| https://openreview.net/pdf?id=S1Ow_e-Rb | |
| PWC | https://paperswithcode.com/paper/how-do-deep-convolutional-neural-networks |
| Repo | |
| Framework | |
Image Blind Denoising With Generative Adversarial Network Based Noise Modeling
| Title | Image Blind Denoising With Generative Adversarial Network Based Noise Modeling |
| Authors | Jingwen Chen, Jiawei Chen, Hongyang Chao, Ming Yang |
| Abstract | In this paper, we consider a typical image blind denoising problem, which is to remove unknown noise from noisy images. As we all know, discriminative learning based methods, such as DnCNN, can achieve state-of-the-art denoising results, but they are not applicable to this problem due to the lack of paired training data. To tackle the barrier, we propose a novel two-step framework. First, a Generative Adversarial Network (GAN) is trained to estimate the noise distribution over the input noisy images and to generate noise samples. Second, the noise patches sampled from the first step are utilized to construct a paired training dataset, which is used, in turn, to train a deep Convolutional Neural Network (CNN) for denoising. Extensive experiments have been done to demonstrate the superiority of our approach in image blind denoising. |
| Tasks | Denoising |
| Published | 2018-06-01 |
| URL | http://openaccess.thecvf.com/content_cvpr_2018/html/Chen_Image_Blind_Denoising_CVPR_2018_paper.html |
| http://openaccess.thecvf.com/content_cvpr_2018/papers/Chen_Image_Blind_Denoising_CVPR_2018_paper.pdf | |
| PWC | https://paperswithcode.com/paper/image-blind-denoising-with-generative |
| Repo | |
| Framework | |
Part-Activated Deep Reinforcement Learning for Action Prediction
| Title | Part-Activated Deep Reinforcement Learning for Action Prediction |
| Authors | Lei Chen, Jiwen Lu, Zhanjie Song, Jie Zhou |
| Abstract | In this paper, we propose a part-activated deep reinforcement learning (PA-DRL) for action prediction. Most existing methods for action prediction utilize the evolution of whole frames to model actions, which cannot avoid the noise of the current action, especially in the early prediction. Moreover, the loss of structural information of human body diminishes the capacity of features to describe actions. To address this, we design a PA-DRL to exploit the structure of the human body by extracting skeleton proposals under a deep reinforcement learning framework. Specifically, we extract features from different parts of the human body individually and activate the action-related parts in features to enhance the representation. Our method not only exploits the structure information of the human body, but also considers the saliency part for expressing actions. We evaluate our method on three popular action prediction datasets: UT-Interaction, BIT-Interaction and UCF101. Our experimental results demonstrate that our method achieves the performance with state-of-the-arts. |
| Tasks | |
| Published | 2018-09-01 |
| URL | http://openaccess.thecvf.com/content_ECCV_2018/html/Lei_Chen_Part-Activated_Deep_Reinforcement_ECCV_2018_paper.html |
| http://openaccess.thecvf.com/content_ECCV_2018/papers/Lei_Chen_Part-Activated_Deep_Reinforcement_ECCV_2018_paper.pdf | |
| PWC | https://paperswithcode.com/paper/part-activated-deep-reinforcement-learning |
| Repo | |
| Framework | |
Building effective deep neural networks one feature at a time
| Title | Building effective deep neural networks one feature at a time |
| Authors | Martin Mundt, Tobias Weis, Kishore Konda, Visvanathan Ramesh |
| Abstract | Successful training of convolutional neural networks is often associated with suffi- ciently deep architectures composed of high amounts of features. These networks typically rely on a variety of regularization and pruning techniques to converge to less redundant states. We introduce a novel bottom-up approach to expand representations in fixed-depth architectures. These architectures start from just a single feature per layer and greedily increase width of individual layers to attain effective representational capacities needed for a specific task. While network growth can rely on a family of metrics, we propose a computationally efficient version based on feature time evolution and demonstrate its potency in determin- ing feature importance and a networks’ effective capacity. We demonstrate how automatically expanded architectures converge to similar topologies that benefit from lesser amount of parameters or improved accuracy and exhibit systematic correspondence in representational complexity with the specified task. In contrast to conventional design patterns with a typical monotonic increase in the amount of features with increased depth, we observe that CNNs perform better when there is more learnable parameters in intermediate, with falloffs to earlier and later layers. |
| Tasks | Feature Importance |
| Published | 2018-01-01 |
| URL | https://openreview.net/forum?id=SkffVjUaW |
| https://openreview.net/pdf?id=SkffVjUaW | |
| PWC | https://paperswithcode.com/paper/building-effective-deep-neural-networks-one |
| Repo | |
| Framework | |
Snap Angle Prediction for 360° Panoramas
| Title | Snap Angle Prediction for 360° Panoramas |
| Authors | Bo Xiong, Kristen Grauman |
| Abstract | 360° panoramas are a rich medium, yet notoriously difficult to visualize in the 2D image plane. We explore how intelligent rotations of a spherical image may enable content-aware projection with fewer perceptible distortions. Whereas existing approaches assume the viewpoint is fixed, intuitively some viewing angles within the sphere preserve high-level objects better than others. To discover the relationship between these optimal emph{snap angles} and the spherical panorama’s content, we develop a reinforcement learning approach for the cubemap projection model. Implemented as a deep recurrent neural network, our method selects a sequence of rotation actions and receives reward for avoiding cube boundaries that overlap with important foreground objects. Our results demonstrate the impact both qualitatively and quantitatively. |
| Tasks | |
| Published | 2018-09-01 |
| URL | http://openaccess.thecvf.com/content_ECCV_2018/html/Bo_Xiong_Snap_Angle_Prediction_ECCV_2018_paper.html |
| http://openaccess.thecvf.com/content_ECCV_2018/papers/Bo_Xiong_Snap_Angle_Prediction_ECCV_2018_paper.pdf | |
| PWC | https://paperswithcode.com/paper/snap-angle-prediction-for-360a-panoramas |
| Repo | |
| Framework | |
Creating a Dataset for Multilingual Fine-grained Emotion-detection Using Gamification-based Annotation
| Title | Creating a Dataset for Multilingual Fine-grained Emotion-detection Using Gamification-based Annotation |
| Authors | Emily {"O}hman, Kaisla Kajava, J{"o}rg Tiedemann, Timo Honkela |
| Abstract | This paper introduces a gamified framework for fine-grained sentiment analysis and emotion detection. We present a flexible tool, \textit{Sentimentator}, that can be used for efficient annotation based on crowd sourcing and a self-perpetuating gold standard. We also present a novel dataset with multi-dimensional annotations of emotions and sentiments in movie subtitles that enables research on sentiment preservation across languages and the creation of robust multilingual emotion detection tools. The tools and datasets are public and open-source and can easily be extended and applied for various purposes. |
| Tasks | Sentiment Analysis |
| Published | 2018-10-01 |
| URL | https://www.aclweb.org/anthology/W18-6205/ |
| https://www.aclweb.org/anthology/W18-6205 | |
| PWC | https://paperswithcode.com/paper/creating-a-dataset-for-multilingual-fine |
| Repo | |
| Framework | |
