October 15, 2019

1928 words 10 mins read

Paper Group NANR 235

Paper Group NANR 235

RefocusGAN: Scene Refocusing using a Single Image. Non-monotone Submodular Maximization in Exponentially Fewer Iterations. ML-LocNet: Improving Object Localization with Multi-view Learning Network. Iterative Language Model Adaptation for Indo-Aryan Language Identification. Towards a Formal Description of NPI-licensing Patterns. Improving Optimizati …

RefocusGAN: Scene Refocusing using a Single Image

Title RefocusGAN: Scene Refocusing using a Single Image
Authors Parikshit Sakurikar, Ishit Mehta, Vineeth N. Balasubramanian, P. J. Narayanan
Abstract Post-capture control of the focus position of an image is a useful photographic tool. Changing the focus of a single image involves the complex task of simultaneously estimating the radiance and the defocus radius of all scene points. We introduce RefocusGAN, a deblur-then-reblur approach to single image refocusing. We train conditional adversarial networks for deblurring and refocusing using wide-aperture images created from light-fields. By appropriately conditioning our networks with a focus measure, an in-focus image and a refocus control parameter, we are able to achieve generic free-form refocusing over a single image.
Tasks Deblurring
Published 2018-09-01
URL http://openaccess.thecvf.com/content_ECCV_2018/html/Parikshit_Sakurikar_Single_Image_Scene_ECCV_2018_paper.html
PDF http://openaccess.thecvf.com/content_ECCV_2018/papers/Parikshit_Sakurikar_Single_Image_Scene_ECCV_2018_paper.pdf
PWC https://paperswithcode.com/paper/refocusgan-scene-refocusing-using-a-single
Repo
Framework

Non-monotone Submodular Maximization in Exponentially Fewer Iterations

Title Non-monotone Submodular Maximization in Exponentially Fewer Iterations
Authors Eric Balkanski, Adam Breuer, Yaron Singer
Abstract In this paper we consider parallelization for applications whose objective can be expressed as maximizing a non-monotone submodular function under a cardinality constraint. Our main result is an algorithm whose approximation is arbitrarily close to 1/2e in O(log^2 n) adaptive rounds, where n is the size of the ground set. This is an exponential speedup in parallel running time over any previously studied algorithm for constrained non-monotone submodular maximization. Beyond its provable guarantees, the algorithm performs well in practice. Specifically, experiments on traffic monitoring and personalized data summarization applications show that the algorithm finds solutions whose values are competitive with state-of-the-art algorithms while running in exponentially fewer parallel iterations.
Tasks Data Summarization
Published 2018-12-01
URL http://papers.nips.cc/paper/7503-non-monotone-submodular-maximization-in-exponentially-fewer-iterations
PDF http://papers.nips.cc/paper/7503-non-monotone-submodular-maximization-in-exponentially-fewer-iterations.pdf
PWC https://paperswithcode.com/paper/non-monotone-submodular-maximization-in
Repo
Framework

ML-LocNet: Improving Object Localization with Multi-view Learning Network

Title ML-LocNet: Improving Object Localization with Multi-view Learning Network
Authors Xiaopeng Zhang, Yang Yang, Jiashi Feng
Abstract This paper addresses Weakly Supervised Object Localization (WSOL) with only image-level supervision. We propose a Multi-view Learning Localization Network (ML-LocNet) by incorporating multi-view learning into a two-phase WSOL model. The multi-view learning would benefit localization due to the complementary relationships among the learned features from different views and the consensus property among the mined instances from each view. In the first phase, the representation is augmented by integrating features learned from multiple views, and in the second phase, the model performs multi-view co-training to enhance localization performance of one view with the help of instances mined from other views, which thus effectively avoids early fitting. ML-LocNet can be easily combined with existing WSOL models to further improve the localization accuracy. Its effectiveness has been proved experimentally. Notably, it achieves 68.6% CorLoc and 49.7% mAP on PASCAL VOC 2007, surpassing the state-of-the-arts by a large margin.
Tasks MULTI-VIEW LEARNING, Object Localization, Weakly-Supervised Object Localization
Published 2018-09-01
URL http://openaccess.thecvf.com/content_ECCV_2018/html/Xiaopeng_Zhang_ML-LocNet_Improving_Object_ECCV_2018_paper.html
PDF http://openaccess.thecvf.com/content_ECCV_2018/papers/Xiaopeng_Zhang_ML-LocNet_Improving_Object_ECCV_2018_paper.pdf
PWC https://paperswithcode.com/paper/ml-locnet-improving-object-localization-with
Repo
Framework

Iterative Language Model Adaptation for Indo-Aryan Language Identification

Title Iterative Language Model Adaptation for Indo-Aryan Language Identification
Authors Tommi Jauhiainen, Heidi Jauhiainen, Krister Lind{'e}n
Abstract This paper presents the experiments and results obtained by the SUKI team in the Indo-Aryan Language Identification shared task of the VarDial 2018 Evaluation Campaign. The shared task was an open one, but we did not use any corpora other than what was distributed by the organizers. A total of eight teams provided results for this shared task. Our submission using a HeLI-method based language identifier with iterative language model adaptation obtained the best results in the shared task with a macro F1-score of 0.958.
Tasks Language Identification, Language Modelling
Published 2018-08-01
URL https://www.aclweb.org/anthology/W18-3907/
PDF https://www.aclweb.org/anthology/W18-3907
PWC https://paperswithcode.com/paper/iterative-language-model-adaptation-for-indo
Repo
Framework

Towards a Formal Description of NPI-licensing Patterns

Title Towards a Formal Description of NPI-licensing Patterns
Authors Mai Ha Vu
Abstract
Tasks
Published 2018-01-01
URL https://www.aclweb.org/anthology/W18-0316/
PDF https://www.aclweb.org/anthology/W18-0316
PWC https://paperswithcode.com/paper/towards-a-formal-description-of-npi-licensing
Repo
Framework

Improving Optimization in Models With Continuous Symmetry Breaking

Title Improving Optimization in Models With Continuous Symmetry Breaking
Authors Robert Bamler, Stephan Mandt
Abstract Many loss functions in representation learning are invariant under a continuous symmetry transformation. For example, the loss function of word embeddings (Mikolov et al., 2013) remains unchanged if we simultaneously rotate all word and context embedding vectors. We show that representation learning models for time series possess an approximate continuous symmetry that leads to slow convergence of gradient descent. We propose a new optimization algorithm that speeds up convergence using ideas from gauge theory in physics. Our algorithm leads to orders of magnitude faster convergence and to more interpretable representations, as we show for dynamic extensions of matrix factorization and word embedding models. We further present an example application of our proposed algorithm that translates modern words into their historic equivalents.
Tasks Representation Learning, Time Series, Word Embeddings
Published 2018-07-01
URL https://icml.cc/Conferences/2018/Schedule?showEvent=2360
PDF http://proceedings.mlr.press/v80/bamler18a/bamler18a.pdf
PWC https://paperswithcode.com/paper/improving-optimization-in-models-with
Repo
Framework

REFUEL: Exploring Sparse Features in Deep Reinforcement Learning for Fast Disease Diagnosis

Title REFUEL: Exploring Sparse Features in Deep Reinforcement Learning for Fast Disease Diagnosis
Authors Yu-Shao Peng, Kai-Fu Tang, Hsuan-Tien Lin, Edward Chang
Abstract This paper proposes REFUEL, a reinforcement learning method with two techniques: {\em reward shaping} and {\em feature rebuilding}, to improve the performance of online symptom checking for disease diagnosis. Reward shaping can guide the search of policy towards better directions. Feature rebuilding can guide the agent to learn correlations between features. Together, they can find symptom queries that can yield positive responses from a patient with high probability. Experimental results justify that the two techniques in REFUEL allows the symptom checker to identify the disease more rapidly and accurately.
Tasks
Published 2018-12-01
URL http://papers.nips.cc/paper/7962-refuel-exploring-sparse-features-in-deep-reinforcement-learning-for-fast-disease-diagnosis
PDF http://papers.nips.cc/paper/7962-refuel-exploring-sparse-features-in-deep-reinforcement-learning-for-fast-disease-diagnosis.pdf
PWC https://paperswithcode.com/paper/refuel-exploring-sparse-features-in-deep
Repo
Framework

Building a Corpus from Handwritten Picture Postcards: Transcription, Annotation and Part-of-Speech Tagging

Title Building a Corpus from Handwritten Picture Postcards: Transcription, Annotation and Part-of-Speech Tagging
Authors Kyoko Sugisaki, Nicolas Wiedmer, Heiko Hausendorf
Abstract
Tasks Optical Character Recognition, Part-Of-Speech Tagging
Published 2018-05-01
URL https://www.aclweb.org/anthology/L18-1038/
PDF https://www.aclweb.org/anthology/L18-1038
PWC https://paperswithcode.com/paper/building-a-corpus-from-handwritten-picture
Repo
Framework

Learning Others’ Intentional Models in Multi-Agent Settings Using Interactive POMDPs

Title Learning Others’ Intentional Models in Multi-Agent Settings Using Interactive POMDPs
Authors Yanlin Han, Piotr Gmytrasiewicz
Abstract Interactive partially observable Markov decision processes (I-POMDPs) provide a principled framework for planning and acting in a partially observable, stochastic and multi-agent environment. It extends POMDPs to multi-agent settings by including models of other agents in the state space and forming a hierarchical belief structure. In order to predict other agents’ actions using I-POMDPs, we propose an approach that effectively uses Bayesian inference and sequential Monte Carlo sampling to learn others’ intentional models which ascribe to them beliefs, preferences and rationality in action selection. Empirical results show that our algorithm accurately learns models of the other agent and has superior performance than methods that use subintentional models. Our approach serves as a generalized Bayesian learning algorithm that learns other agents’ beliefs, strategy levels, and transition, observation and reward functions. It also effectively mitigates the belief space complexity due to the nested belief hierarchy.
Tasks Bayesian Inference
Published 2018-12-01
URL http://papers.nips.cc/paper/7806-learning-others-intentional-models-in-multi-agent-settings-using-interactive-pomdps
PDF http://papers.nips.cc/paper/7806-learning-others-intentional-models-in-multi-agent-settings-using-interactive-pomdps.pdf
PWC https://paperswithcode.com/paper/learning-others-intentional-models-in-multi
Repo
Framework

The First Komi-Zyrian Universal Dependencies Treebanks

Title The First Komi-Zyrian Universal Dependencies Treebanks
Authors Niko Partanen, Rogier Blokland, KyungTae Lim, Thierry Poibeau, Michael Rießler
Abstract
Tasks Dependency Parsing
Published 2018-11-01
URL https://www.aclweb.org/anthology/papers/W18-6015/w18-6015
PDF https://www.aclweb.org/anthology/W18-6015
PWC https://paperswithcode.com/paper/the-first-komi-zyrian-universal-dependencies
Repo
Framework

Expert Evaluation of a Spoken Dialogue System in a Clinical Operating Room

Title Expert Evaluation of a Spoken Dialogue System in a Clinical Operating Room
Authors Juliana Miehle, Nadine Gerstenlauer, Daniel Ostler, Hubertus Feu{\ss}ner, Wolfgang Minker, Stefan Ultes
Abstract
Tasks Spoken Dialogue Systems
Published 2018-05-01
URL https://www.aclweb.org/anthology/L18-1118/
PDF https://www.aclweb.org/anthology/L18-1118
PWC https://paperswithcode.com/paper/expert-evaluation-of-a-spoken-dialogue-system
Repo
Framework

Accelerating Dynamic Programs via Nested Benders Decomposition with Application to Multi-Person Pose Estimation

Title Accelerating Dynamic Programs via Nested Benders Decomposition with Application to Multi-Person Pose Estimation
Authors Shaofei Wang, Alexander Ihler, Konrad Kording, Julian Yarkony
Abstract We present a novel approach to solve dynamic programs (DP), which are frequent in computer vision, on tree-structured graphs with exponential node state space. Typical DP approaches have to enumerate the joint state space of two adjacent nodes on every edge of the tree to compute the optimal messages. Here we propose an algorithm based on Nested Benders Decomposition (NBD) which iteratively lower-bounds the message on every edge and promises to be far more efficient. We apply our NBD algorithm along with a novel Minimum Weight Set Packing (MWSP) formulation to a multi-person pose estimation problem. While our algorithm is provably optimal at termination it operates in linear time for practical DP problems, gaining up to 500x speed up over traditional DP algorithm which have polynomial complexity.
Tasks Multi-Person Pose Estimation, Pose Estimation
Published 2018-09-01
URL http://openaccess.thecvf.com/content_ECCV_2018/html/Shaofei_Wang_Accelerating_Dynamic_Programs_ECCV_2018_paper.html
PDF http://openaccess.thecvf.com/content_ECCV_2018/papers/Shaofei_Wang_Accelerating_Dynamic_Programs_ECCV_2018_paper.pdf
PWC https://paperswithcode.com/paper/accelerating-dynamic-programs-via-nested
Repo
Framework

Learning to Listen: Critically Considering the Role of AI in Human Storytelling and Character Creation

Title Learning to Listen: Critically Considering the Role of AI in Human Storytelling and Character Creation
Authors Anna Kasunic, Geoff Kaufman
Abstract In this opinion piece, we argue that there is a need for alternative design directions to complement existing AI efforts in narrative and character generation and algorithm development. To make our argument, we a) outline the predominant roles and goals of AI research in storytelling; b) present existing discourse on the benefits and harms of narratives; and c) highlight the pain points in character creation revealed by semi-structured interviews we conducted with 14 individuals deeply involved in some form of character creation. We conclude by proffering several specific design avenues that we believe can seed fruitful research collaborations. In our vision, AI collaborates with humans during creative processes and narrative generation, helps amplify voices and perspectives that are currently marginalized or misrepresented, and engenders experiences of narrative that support spectatorship and listening roles.
Tasks
Published 2018-06-01
URL https://www.aclweb.org/anthology/W18-1501/
PDF https://www.aclweb.org/anthology/W18-1501
PWC https://paperswithcode.com/paper/learning-to-listen-critically-considering-the
Repo
Framework

MULTI-MODAL EMOTION RECOGNITION ON IEMOCAP WITH NEURAL NETWORKS.

Title MULTI-MODAL EMOTION RECOGNITION ON IEMOCAP WITH NEURAL NETWORKS.
Authors Samarth Tripathi, Homayoon Beigi
Abstract Emotion recognition has become an important field of re- search in Human Computer Interactions and there is a grow- ing need for automatic emotion recognition systems. One of the directions the research is heading is the use of Neural Networks which are adept at estimating complex functions that depend on a large number and diverse source of input data. In this paper we attempt to exploit this effectiveness of Neural networks to enable us to perform multimodal Emotion recognition on IEMOCAP dataset using data from Speech, Text, and Motions captured from face expressions, rotation and hand movements. Prior research has concentrated on Emotion detection from Speech on the IEMOCAP dataset, but our approach uses the multiple modes of data offered by IEMOCAP for a more robust and accurate emotion detection
Tasks Emotion Recognition, Multimodal Emotion Recognition
Published 2018-11-12
URL https://www.paperswithcode.com/submit-paper
PDF https://www.paperswithcode.com/submit-paper
PWC https://paperswithcode.com/paper/multi-modal-emotion-recognition-on-iemocap-1
Repo
Framework

An Arabic Morphological Analyzer and Generator with Copious Features

Title An Arabic Morphological Analyzer and Generator with Copious Features
Authors Dima Taji, Salam Khalifa, Ossama Obeid, Fadhl Eryani, Nizar Habash
Abstract We introduce CALIMA-Star, a very rich Arabic morphological analyzer and generator that provides functional and form-based morphological features as well as built-in tokenization, phonological representation, lexical rationality and much more. This tool includes a fast engine that can be easily integrated into other systems, as well as an easy-to-use API and a web interface. CALIMA-Star also supports morphological reinflection. We evaluate CALIMA-Star against four commonly used analyzers for Arabic in terms of speed and morphological content.
Tasks Tokenization
Published 2018-10-01
URL https://www.aclweb.org/anthology/W18-5816/
PDF https://www.aclweb.org/anthology/W18-5816
PWC https://paperswithcode.com/paper/an-arabic-morphological-analyzer-and
Repo
Framework
comments powered by Disqus