Paper Group AWR 130
RDF2PT: Generating Brazilian Portuguese Texts from RDF Data. Stochastic Answer Networks for Natural Language Inference. Improving Context Modelling in Multimodal Dialogue Generation. Improving Conditional Sequence Generative Adversarial Networks by Stepwise Evaluation. A Factor Graph Approach to Automated Design of Bayesian Signal Processing Algori …
RDF2PT: Generating Brazilian Portuguese Texts from RDF Data
Title | RDF2PT: Generating Brazilian Portuguese Texts from RDF Data |
Authors | Diego Moussallem, Thiago Castro Ferreira, Marcos Zampieri, Maria Claudia Cavalcanti, Geraldo Xexéo, Mariana Neves, Axel-Cyrille Ngonga Ngomo |
Abstract | The generation of natural language from Resource Description Framework (RDF) data has recently gained significant attention due to the continuous growth of Linked Data. A number of these approaches generate natural language in languages other than English, however, no work has been proposed to generate Brazilian Portuguese texts out of RDF. We address this research gap by presenting RDF2PT, an approach that verbalizes RDF data to Brazilian Portuguese language. We evaluated RDF2PT in an open questionnaire with 44 native speakers divided into experts and non-experts. Our results suggest that RDF2PT is able to generate text which is similar to that generated by humans and can hence be easily understood. |
Tasks | |
Published | 2018-02-22 |
URL | http://arxiv.org/abs/1802.08150v1 |
http://arxiv.org/pdf/1802.08150v1.pdf | |
PWC | https://paperswithcode.com/paper/rdf2pt-generating-brazilian-portuguese-texts |
Repo | https://github.com/dice-group/RDF2PT |
Framework | none |
Stochastic Answer Networks for Natural Language Inference
Title | Stochastic Answer Networks for Natural Language Inference |
Authors | Xiaodong Liu, Kevin Duh, Jianfeng Gao |
Abstract | We propose a stochastic answer network (SAN) to explore multi-step inference strategies in Natural Language Inference. Rather than directly predicting the results given the inputs, the model maintains a state and iteratively refines its predictions. Our experiments show that SAN achieves the state-of-the-art results on three benchmarks: Stanford Natural Language Inference (SNLI) dataset, MultiGenre Natural Language Inference (MultiNLI) dataset and Quora Question Pairs dataset. |
Tasks | Natural Language Inference |
Published | 2018-04-21 |
URL | http://arxiv.org/abs/1804.07888v2 |
http://arxiv.org/pdf/1804.07888v2.pdf | |
PWC | https://paperswithcode.com/paper/stochastic-answer-networks-for-natural |
Repo | https://github.com/kevinduh/san_mrc |
Framework | pytorch |
Improving Context Modelling in Multimodal Dialogue Generation
Title | Improving Context Modelling in Multimodal Dialogue Generation |
Authors | Shubham Agarwal, Ondrej Dusek, Ioannis Konstas, Verena Rieser |
Abstract | In this work, we investigate the task of textual response generation in a multimodal task-oriented dialogue system. Our work is based on the recently released Multimodal Dialogue (MMD) dataset (Saha et al., 2017) in the fashion domain. We introduce a multimodal extension to the Hierarchical Recurrent Encoder-Decoder (HRED) model and show that this extension outperforms strong baselines in terms of text-based similarity metrics. We also showcase the shortcomings of current vision and language models by performing an error analysis on our system’s output. |
Tasks | Dialogue Generation |
Published | 2018-10-20 |
URL | http://arxiv.org/abs/1810.11955v1 |
http://arxiv.org/pdf/1810.11955v1.pdf | |
PWC | https://paperswithcode.com/paper/improving-context-modelling-in-multimodal |
Repo | https://github.com/shubhamagarwal92/mmd |
Framework | pytorch |
Improving Conditional Sequence Generative Adversarial Networks by Stepwise Evaluation
Title | Improving Conditional Sequence Generative Adversarial Networks by Stepwise Evaluation |
Authors | Yi-Lin Tuan, Hung-Yi Lee |
Abstract | Sequence generative adversarial networks (SeqGAN) have been used to improve conditional sequence generation tasks, for example, chit-chat dialogue generation. To stabilize the training of SeqGAN, Monte Carlo tree search (MCTS) or reward at every generation step (REGS) is used to evaluate the goodness of a generated subsequence. MCTS is computationally intensive, but the performance of REGS is worse than MCTS. In this paper, we propose stepwise GAN (StepGAN), in which the discriminator is modified to automatically assign scores quantifying the goodness of each subsequence at every generation step. StepGAN has significantly less computational costs than MCTS. We demonstrate that StepGAN outperforms previous GAN-based methods on both synthetic experiment and chit-chat dialogue generation. |
Tasks | Dialogue Generation |
Published | 2018-08-16 |
URL | http://arxiv.org/abs/1808.05599v2 |
http://arxiv.org/pdf/1808.05599v2.pdf | |
PWC | https://paperswithcode.com/paper/improving-conditional-sequence-generative |
Repo | https://github.com/Pascalson/Conditional-Seq-GANs |
Framework | tf |
A Factor Graph Approach to Automated Design of Bayesian Signal Processing Algorithms
Title | A Factor Graph Approach to Automated Design of Bayesian Signal Processing Algorithms |
Authors | Marco Cox, Thijs van de Laar, Bert de Vries |
Abstract | The benefits of automating design cycles for Bayesian inference-based algorithms are becoming increasingly recognized by the machine learning community. As a result, interest in probabilistic programming frameworks has much increased over the past few years. This paper explores a specific probabilistic programming paradigm, namely message passing in Forney-style factor graphs (FFGs), in the context of automated design of efficient Bayesian signal processing algorithms. To this end, we developed “ForneyLab” (https://github.com/biaslab/ForneyLab.jl) as a Julia toolbox for message passing-based inference in FFGs. We show by example how ForneyLab enables automatic derivation of Bayesian signal processing algorithms, including algorithms for parameter estimation and model comparison. Crucially, due to the modular makeup of the FFG framework, both the model specification and inference methods are readily extensible in ForneyLab. In order to test this framework, we compared variational message passing as implemented by ForneyLab with automatic differentiation variational inference (ADVI) and Monte Carlo methods as implemented by state-of-the-art tools “Edward” and “Stan”. In terms of performance, extensibility and stability issues, ForneyLab appears to enjoy an edge relative to its competitors for automated inference in state-space models. |
Tasks | Bayesian Inference, Probabilistic Programming |
Published | 2018-11-08 |
URL | http://arxiv.org/abs/1811.03407v1 |
http://arxiv.org/pdf/1811.03407v1.pdf | |
PWC | https://paperswithcode.com/paper/a-factor-graph-approach-to-automated-design |
Repo | https://github.com/biaslab/ForneyLab.jl |
Framework | none |
Simplifying Neural Machine Translation with Addition-Subtraction Twin-Gated Recurrent Networks
Title | Simplifying Neural Machine Translation with Addition-Subtraction Twin-Gated Recurrent Networks |
Authors | Biao Zhang, Deyi Xiong, Jinsong Su, Qian Lin, Huiji Zhang |
Abstract | In this paper, we propose an additionsubtraction twin-gated recurrent network (ATR) to simplify neural machine translation. The recurrent units of ATR are heavily simplified to have the smallest number of weight matrices among units of all existing gated RNNs. With the simple addition and subtraction operation, we introduce a twin-gated mechanism to build input and forget gates which are highly correlated. Despite this simplification, the essential non-linearities and capability of modeling long-distance dependencies are preserved. Additionally, the proposed ATR is more transparent than LSTM/GRU due to the simplification. Forward self-attention can be easily established in ATR, which makes the proposed network interpretable. Experiments on WMT14 translation tasks demonstrate that ATR-based neural machine translation can yield competitive performance on English- German and English-French language pairs in terms of both translation quality and speed. Further experiments on NIST Chinese-English translation, natural language inference and Chinese word segmentation verify the generality and applicability of ATR on different natural language processing tasks. |
Tasks | Chinese Word Segmentation, Machine Translation, Natural Language Inference |
Published | 2018-10-30 |
URL | http://arxiv.org/abs/1810.12546v1 |
http://arxiv.org/pdf/1810.12546v1.pdf | |
PWC | https://paperswithcode.com/paper/simplifying-neural-machine-translation-with |
Repo | https://github.com/bzhangGo/zero |
Framework | tf |
FPD-M-net: Fingerprint Image Denoising and Inpainting Using M-Net Based Convolutional Neural Networks
Title | FPD-M-net: Fingerprint Image Denoising and Inpainting Using M-Net Based Convolutional Neural Networks |
Authors | Sukesh Adiga V, Jayanthi Sivaswamy |
Abstract | Fingerprint is a common biometric used for authentication and verification of an individual. These images are degraded when fingers are wet, dirty, dry or wounded and due to the failure of the sensors, etc. The extraction of the fingerprint from a degraded image requires denoising and inpainting. We propose to address these problems with an end-to-end trainable Convolutional Neural Network based architecture called FPD-M-net, by posing the fingerprint denoising and inpainting problem as a segmentation (foreground) task. Our architecture is based on the M-net with a change: structure similarity loss function, used for better extraction of the fingerprint from the noisy background. Our method outperforms the baseline method and achieves an overall 3rd rank in the Chalearn LAP Inpainting Competition Track 3 - Fingerprint Denoising and Inpainting, ECCV 2018 |
Tasks | Denoising, Image Denoising |
Published | 2018-12-26 |
URL | http://arxiv.org/abs/1812.10191v2 |
http://arxiv.org/pdf/1812.10191v2.pdf | |
PWC | https://paperswithcode.com/paper/fpd-m-net-fingerprint-image-denoising-and |
Repo | https://github.com/adigasu/FDPMNet |
Framework | tf |
BranchGAN: Branched Generative Adversarial Networks for Scale-Disentangled Learning and Synthesis of Images
Title | BranchGAN: Branched Generative Adversarial Networks for Scale-Disentangled Learning and Synthesis of Images |
Authors | Zili Yi, Zhiqin Chen, Hao Cai, Xin Huang, Minglun Gong, Hao Zhang |
Abstract | We introduce BranchGAN, a novel training method that enables unconditioned generative adversarial networks (GANs) to learn image manifolds at multiple scales. The key novel feature of BranchGAN is that it is trained in multiple branches, progressively covering both the breadth and depth of the network, as resolutions of the training images increase to reveal finer-scale features. Specifically, each noise vector, as input to the generator network, is explicitly split into several sub-vectors, each corresponding to, and is trained to learn, image representations at a particular scale. During training, we progressively de-freeze the sub-vectors, one at a time, as a new set of higher-resolution images is employed for training and more network layers are added. A consequence of such an explicit sub-vector designation is that we can directly manipulate and even combine latent (sub-vector) codes which model different feature scales. Experiments demonstrate the effectiveness of our training method in scale-disentangled learning of image manifolds and synthesis, without any extra labels and without compromising quality of the synthesized high-resolution images. We further demonstrate three applications enabled or improved by BranchGAN. |
Tasks | |
Published | 2018-03-22 |
URL | http://arxiv.org/abs/1803.08467v2 |
http://arxiv.org/pdf/1803.08467v2.pdf | |
PWC | https://paperswithcode.com/paper/branchgan-branched-generative-adversarial |
Repo | https://github.com/duxingren14/BranchGAN |
Framework | tf |
Generalized Robust Bayesian Committee Machine for Large-scale Gaussian Process Regression
Title | Generalized Robust Bayesian Committee Machine for Large-scale Gaussian Process Regression |
Authors | Haitao Liu, Jianfei Cai, Yi Wang, Yew-Soon Ong |
Abstract | In order to scale standard Gaussian process (GP) regression to large-scale datasets, aggregation models employ factorized training process and then combine predictions from distributed experts. The state-of-the-art aggregation models, however, either provide inconsistent predictions or require time-consuming aggregation process. We first prove the inconsistency of typical aggregations using disjoint or random data partition, and then present a consistent yet efficient aggregation model for large-scale GP. The proposed model inherits the advantages of aggregations, e.g., closed-form inference and aggregation, parallelization and distributed computing. Furthermore, theoretical and empirical analyses reveal that the new aggregation model performs better due to the consistent predictions that converge to the true underlying function when the training size approaches infinity. |
Tasks | |
Published | 2018-06-03 |
URL | http://arxiv.org/abs/1806.00720v1 |
http://arxiv.org/pdf/1806.00720v1.pdf | |
PWC | https://paperswithcode.com/paper/generalized-robust-bayesian-committee-machine |
Repo | https://github.com/LiuHaiTao01/GRBCM |
Framework | none |
Temporal Interpolation via Motion Field Prediction
Title | Temporal Interpolation via Motion Field Prediction |
Authors | Lin Zhang, Neerav Karani, Christine Tanner, Ender Konukoglu |
Abstract | Navigated 2D multi-slice dynamic Magnetic Resonance (MR) imaging enables high contrast 4D MR imaging during free breathing and provides in-vivo observations for treatment planning and guidance. Navigator slices are vital for retrospective stacking of 2D data slices in this method. However, they also prolong the acquisition sessions. Temporal interpolation of navigator slices an be used to reduce the number of navigator acquisitions without degrading specificity in stacking. In this work, we propose a convolutional neural network (CNN) based method for temporal interpolation via motion field prediction. The proposed formulation incorporates the prior knowledge that a motion field underlies changes in the image intensities over time. Previous approaches that interpolate directly in the intensity space are prone to produce blurry images or even remove structures in the images. Our method avoids such problems and faithfully preserves the information in the image. Further, an important advantage of our formulation is that it provides an unsupervised estimation of bi-directional motion fields. We show that these motion fields can be used to halve the number of registrations required during 4D reconstruction, thus substantially reducing the reconstruction time. |
Tasks | |
Published | 2018-04-12 |
URL | http://arxiv.org/abs/1804.04440v1 |
http://arxiv.org/pdf/1804.04440v1.pdf | |
PWC | https://paperswithcode.com/paper/temporal-interpolation-via-motion-field |
Repo | https://github.com/linz94/mfin-cycle |
Framework | tf |
Meta Architecture Search
Title | Meta Architecture Search |
Authors | Albert Shaw, Wei Wei, Weiyang Liu, Le Song, Bo Dai |
Abstract | Neural Architecture Search (NAS) has been quite successful in constructing state-of-the-art models on a variety of tasks. Unfortunately, the computational cost can make it difficult to scale. In this paper, we make the first attempt to study Meta Architecture Search which aims at learning a task-agnostic representation that can be used to speed up the process of architecture search on a large number of tasks. We propose the Bayesian Meta Architecture SEarch (BASE) framework which takes advantage of a Bayesian formulation of the architecture search problem to learn over an entire set of tasks simultaneously. We show that on Imagenet classification, we can find a model that achieves 25.7% top-1 error and 8.1% top-5 error by adapting the architecture in less than an hour from an 8 GPU days pretrained meta-network. By learning a good prior for NAS, our method dramatically decreases the required computation cost while achieving comparable performance to current state-of-the-art methods - even finding competitive models for unseen datasets with very quick adaptation. We believe our framework will open up new possibilities for efficient and massively scalable architecture search research across multiple tasks. |
Tasks | Bayesian Inference, Few-Shot Learning, Meta-Learning, Neural Architecture Search |
Published | 2018-12-22 |
URL | https://arxiv.org/abs/1812.09584v2 |
https://arxiv.org/pdf/1812.09584v2.pdf | |
PWC | https://paperswithcode.com/paper/bayesian-meta-network-architecture-learning |
Repo | https://github.com/ashaw596/meta_architecture_search |
Framework | pytorch |
Double Embeddings and CNN-based Sequence Labeling for Aspect Extraction
Title | Double Embeddings and CNN-based Sequence Labeling for Aspect Extraction |
Authors | Hu Xu, Bing Liu, Lei Shu, Philip S. Yu |
Abstract | One key task of fine-grained sentiment analysis of product reviews is to extract product aspects or features that users have expressed opinions on. This paper focuses on supervised aspect extraction using deep learning. Unlike other highly sophisticated supervised deep learning models, this paper proposes a novel and yet simple CNN model employing two types of pre-trained embeddings for aspect extraction: general-purpose embeddings and domain-specific embeddings. Without using any additional supervision, this model achieves surprisingly good results, outperforming state-of-the-art sophisticated existing methods. To our knowledge, this paper is the first to report such double embeddings based CNN model for aspect extraction and achieve very good results. |
Tasks | Aspect-Based Sentiment Analysis, Aspect Extraction, Sentiment Analysis |
Published | 2018-05-11 |
URL | http://arxiv.org/abs/1805.04601v1 |
http://arxiv.org/pdf/1805.04601v1.pdf | |
PWC | https://paperswithcode.com/paper/double-embeddings-and-cnn-based-sequence |
Repo | https://github.com/howardhsu/DE-CNN |
Framework | pytorch |
DisguiseNet : A Contrastive Approach for Disguised Face Verification in the Wild
Title | DisguiseNet : A Contrastive Approach for Disguised Face Verification in the Wild |
Authors | Skand Vishwanath Peri, Abhinav Dhall |
Abstract | This paper describes our approach for the Disguised Faces in the Wild (DFW) 2018 challenge. The task here is to verify the identity of a person among disguised and impostors images. Given the importance of the task of face verification it is essential to compare methods across a common platform. Our approach is based on VGG-face architecture paired with Contrastive loss based on cosine distance metric. For augmenting the data set, we source more data from the internet. The experiments show the effectiveness of the approach on the DFW data. We show that adding extra data to the DFW dataset with noisy labels also helps in increasing the generalization performance of the network. The proposed network achieves 27.13% absolute increase in accuracy over the DFW baseline. |
Tasks | Disguised Face Verification, Face Verification |
Published | 2018-04-25 |
URL | http://arxiv.org/abs/1804.09669v2 |
http://arxiv.org/pdf/1804.09669v2.pdf | |
PWC | https://paperswithcode.com/paper/disguisenet-a-contrastive-approach-for |
Repo | https://github.com/pvskand/DisguiseNet |
Framework | tf |
Distributed Deep Reinforcement Learning: Learn how to play Atari games in 21 minutes
Title | Distributed Deep Reinforcement Learning: Learn how to play Atari games in 21 minutes |
Authors | Igor Adamski, Robert Adamski, Tomasz Grel, Adam Jędrych, Kamil Kaczmarek, Henryk Michalewski |
Abstract | We present a study in Distributed Deep Reinforcement Learning (DDRL) focused on scalability of a state-of-the-art Deep Reinforcement Learning algorithm known as Batch Asynchronous Advantage ActorCritic (BA3C). We show that using the Adam optimization algorithm with a batch size of up to 2048 is a viable choice for carrying out large scale machine learning computations. This, combined with careful reexamination of the optimizer’s hyperparameters, using synchronous training on the node level (while keeping the local, single node part of the algorithm asynchronous) and minimizing the memory footprint of the model, allowed us to achieve linear scaling for up to 64 CPU nodes. This corresponds to a training time of 21 minutes on 768 CPU cores, as opposed to 10 hours when using a single node with 24 cores achieved by a baseline single-node implementation. |
Tasks | Atari Games |
Published | 2018-01-09 |
URL | http://arxiv.org/abs/1801.02852v2 |
http://arxiv.org/pdf/1801.02852v2.pdf | |
PWC | https://paperswithcode.com/paper/distributed-deep-reinforcement-learning-learn |
Repo | https://github.com/deepsense-ai/Distributed-BA3C |
Framework | tf |
DocFace: Matching ID Document Photos to Selfies
Title | DocFace: Matching ID Document Photos to Selfies |
Authors | Yichun Shi, Anil K. Jain |
Abstract | Numerous activities in our daily life, including transactions, access to services and transportation, require us to verify who we are by showing our ID documents containing face images, e.g. passports and driver licenses. An automatic system for matching ID document photos to live face images in real time with high accuracy would speedup the verification process and remove the burden on human operators. In this paper, by employing the transfer learning technique, we propose a new method, DocFace, to train a domain-specific network for ID document photo matching without a large dataset. Compared with the baseline of applying existing methods for general face recognition to this problem, our method achieves considerable improvement. A cross validation on an ID-Selfie dataset shows that DocFace improves the TAR from 61.14% to 92.77% at FAR=0.1%. Experimental results also indicate that given more training data, a viable system for automatic ID document photo matching can be developed and deployed. |
Tasks | Face Recognition, Transfer Learning |
Published | 2018-05-06 |
URL | http://arxiv.org/abs/1805.02283v1 |
http://arxiv.org/pdf/1805.02283v1.pdf | |
PWC | https://paperswithcode.com/paper/docface-matching-id-document-photos-to |
Repo | https://github.com/seasonSH/DocFace |
Framework | tf |