PyTorchKR Papers

최신 논문

총 9,642편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1801.05119

Variational Recurrent Neural Machine Translation

Partially inspired by successful applications of variational recurrent neural networks, we propose a novel variational recurrent neural machine translation (VRNMT) model in this paper. Different from the variational NMT, VRNMT introduces a series of latent random variables to model the translation procedure of a sentence in a generative way, instead of a single latent variable. Specifically, the latent random variables are included into the hidden states of the NMT decoder with elements from the variational autoencoder. In this way, these variables are recurrently generated, which enables them to further capture strong and complex dependencies among the output translations at different timesteps. In order to deal with the challenges in performing efficient posterior inference and large-scale training during the incorporation of latent variables, we build a neural posterior approximator, and equip it with a reparameterization technique to estimate the variational lower bound. Experiments on Chinese-English and English-German translation tasks demonstrate that the proposed model achieves significant improvements over both the conventional and variational NMT models.

arXiv:1709.03009공식 코드 · Papers with Code

How to Train a CAT: Learning Canonical Appearance Transformations for Direct Visual Localization Under Illumination Change

Direct visual localization has recently enjoyed a resurgence in popularity with the increasing availability of cheap mobile computing power. The competitive accuracy and robustness of these algorithms compared to state-of-the-art feature-based methods, as well as their natural ability to yield dense maps, makes them an appealing choice for a variety of mobile robotics applications. However, direct methods remain brittle in the face of appearance change due to their underlying assumption of photometric consistency, which is commonly violated in practice. In this paper, we propose to mitigate this problem by training deep convolutional encoder-decoder models to transform images of a scene such that they correspond to a previously-seen canonical appearance. We validate our method in multiple environments and illumination conditions using high-fidelity synthetic RGB-D datasets, and integrate the trained models into a direct visual localization pipeline, yielding improvements in visual odometry (VO) accuracy through time-varying illumination conditions, as well as improved metric relocalization performance under illumination change, where conventional methods normally fail. We further provide a preliminary investigation of transfer learning from synthetic to real environments in a localization context. An open-source implementation of our method using PyTorch is available at https://github.com/utiasSTARS/cat-net.

arXiv:1711.06439

Towards dense volumetric pancreas segmentation in CT using 3D fully convolutional networks

Pancreas segmentation in computed tomography imaging has been historically difficult for automated methods because of the large shape and size variations between patients. In this work, we describe a custom-build 3D fully convolutional network (FCN) that can process a 3D image including the whole pancreas and produce an automatic segmentation. We investigate two variations of the 3D FCN architecture; one with concatenation and one with summation skip connections to the decoder part of the network. We evaluate our methods on a dataset from a clinical trial with gastric cancer patients, including 147 contrast enhanced abdominal CT scans acquired in the portal venous phase. Using the summation architecture, we achieve an average Dice score of 89.7 $\pm$ 3.8 (range [79.8, 94.8]) % in testing, achieving the new state-of-the-art performance in pancreas segmentation on this dataset.

arXiv:1801.10109

Trajectory-based Radical Analysis Network for Online Handwritten Chinese Character Recognition

Recently, great progress has been made for online handwritten Chinese character recognition due to the emergence of deep learning techniques. However, previous research mostly treated each Chinese character as one class without explicitly considering its inherent structure, namely the radical components with complicated geometry. In this study, we propose a novel trajectory-based radical analysis network (TRAN) to firstly identify radicals and analyze two-dimensional structures among radicals simultaneously, then recognize Chinese characters by generating captions of them based on the analysis of their internal radicals. The proposed TRAN employs recurrent neural networks (RNNs) as both an encoder and a decoder. The RNN encoder makes full use of online information by directly transforming handwriting trajectory into high-level features. The RNN decoder aims at generating the caption by detecting radicals and spatial structures through an attention model. The manner of treating a Chinese character as a two-dimensional composition of radicals can reduce the size of vocabulary and enable TRAN to possess the capability of recognizing unseen Chinese character classes, only if the corresponding radicals have been seen. Evaluated on CASIA-OLHWDB database, the proposed approach significantly outperforms the state-of-the-art whole-character modeling approach with a relative character error rate (CER) reduction of 10%. Meanwhile, for the case of recognition of 500 unseen Chinese characters, TRAN can achieve a character accuracy of about 60% while the traditional whole-character method has no capability to handle them.

arXiv:1801.06879

Bayesian Deep Convolutional Encoder-Decoder Networks for Surrogate Modeling and Uncertainty Quantification

We are interested in the development of surrogate models for uncertainty quantification and propagation in problems governed by stochastic PDEs using a deep convolutional encoder-decoder network in a similar fashion to approaches considered in deep learning for image-to-image regression tasks. Since normal neural networks are data intensive and cannot provide predictive uncertainty, we propose a Bayesian approach to convolutional neural nets. A recently introduced variational gradient descent algorithm based on Stein's method is scaled to deep convolutional networks to perform approximate Bayesian inference on millions of uncertain network parameters. This approach achieves state of the art performance in terms of predictive accuracy and uncertainty quantification in comparison to other approaches in Bayesian neural networks as well as techniques that include Gaussian processes and ensemble methods even when the training data size is relatively small. To evaluate the performance of this approach, we consider standard uncertainty quantification benchmark problems including flow in heterogeneous media defined in terms of limited data-driven permeability realizations. The performance of the surrogate model developed is very good even though there is no underlying structure shared between the input (permeability) and output (flow/pressure) fields as is often the case in the image-to-image regression models used in computer vision problems. Studies are performed with an underlying stochastic input dimensionality up to $4,225$ where most other uncertainty quantification methods fail. Uncertainty propagation tasks are considered and the predictive output Bayesian statistics are compared to those obtained with Monte Carlo estimates.

arXiv:1801.08301

Class label autoencoder for zero-shot learning

Existing zero-shot learning (ZSL) methods usually learn a projection function between a feature space and a semantic embedding space(text or attribute space) in the training seen classes or testing unseen classes. However, the projection function cannot be used between the feature space and multi-semantic embedding spaces, which have the diversity characteristic for describing the different semantic information of the same class. To deal with this issue, we present a novel method to ZSL based on learning class label autoencoder (CLA). CLA can not only build a uniform framework for adapting to multi-semantic embedding spaces, but also construct the encoder-decoder mechanism for constraining the bidirectional projection between the feature space and the class label space. Moreover, CLA can jointly consider the relationship of feature classes and the relevance of the semantic classes for improving zero-shot classification. The CLA solution can provide both unseen class labels and the relation of the different classes representation(feature or semantic information) that can encode the intrinsic structure of classes. Extensive experiments demonstrate the CLA outperforms state-of-art methods on four benchmark datasets, which are AwA, CUB, Dogs and ImNet-2.

arXiv:1703.10893

Audio-Visual Speech Enhancement Using Multimodal Deep Convolutional Neural Networks

Speech enhancement (SE) aims to reduce noise in speech signals. Most SE techniques focus only on addressing audio information. In this work, inspired by multimodal learning, which utilizes data from different modalities, and the recent success of convolutional neural networks (CNNs) in SE, we propose an audio-visual deep CNNs (AVDCNN) SE model, which incorporates audio and visual streams into a unified network model. We also propose a multi-task learning framework for reconstructing audio and visual signals at the output layer. Precisely speaking, the proposed AVDCNN model is structured as an audio-visual encoder-decoder network, in which audio and visual data are first processed using individual CNNs, and then fused into a joint network to generate enhanced speech (the primary task) and reconstructed images (the secondary task) at the output layer. The model is trained in an end-to-end manner, and parameters are jointly learned through back-propagation. We evaluate enhanced speech using five instrumental criteria. Results show that the AVDCNN model yields a notably superior performance compared with an audio-only CNN-based SE model and two conventional SE approaches, confirming the effectiveness of integrating visual information into the SE process. In addition, the AVDCNN model also outperforms an existing audio-visual SE model, confirming its capability of effectively combining audio and visual information in SE.

arXiv:1801.09555

DeepLung: Deep 3D Dual Path Nets for Automated Pulmonary Nodule Detection and Classification

In this work, we present a fully automated lung computed tomography (CT) cancer diagnosis system, DeepLung. DeepLung consists of two components, nodule detection (identifying the locations of candidate nodules) and classification (classifying candidate nodules into benign or malignant). Considering the 3D nature of lung CT data and the compactness of dual path networks (DPN), two deep 3D DPN are designed for nodule detection and classification respectively. Specifically, a 3D Faster Regions with Convolutional Neural Net (R-CNN) is designed for nodule detection with 3D dual path blocks and a U-net-like encoder-decoder structure to effectively learn nodule features. For nodule classification, gradient boosting machine (GBM) with 3D dual path network features is proposed. The nodule classification subnetwork was validated on a public dataset from LIDC-IDRI, on which it achieved better performance than state-of-the-art approaches and surpassed the performance of experienced doctors based on image modality. Within the DeepLung system, candidate nodules are detected first by the nodule detection subnetwork, and nodule diagnosis is conducted by the classification subnetwork. Extensive experimental results demonstrate that DeepLung has performance comparable to experienced doctors both for the nodule-level and patient-level diagnosis on the LIDC-IDRI dataset.\footnote{https://github.com/uci-cbcl/DeepLung.git}

arXiv:1801.08831공식 코드 · Papers with Code

A Multilayer Convolutional Encoder-Decoder Neural Network for Grammatical Error Correction

We improve automatic correction of grammatical, orthographic, and collocation errors in text using a multilayer convolutional encoder-decoder neural network. The network is initialized with embeddings that make use of character N-gram information to better suit this task. When evaluated on common benchmark test data sets (CoNLL-2014 and JFLEG), our model substantially outperforms all prior neural approaches on this task as well as strong statistical machine translation-based systems with neural and task-specific features trained on the same data. Our analysis shows the superiority of convolutional neural networks over recurrent neural networks such as long short-term memory (LSTM) networks in capturing the local context via attention, and thereby improving the coverage in correcting grammatical errors. By ensembling multiple models, and incorporating an N-gram language model and edit features via rescoring, our novel method becomes the first neural approach to outperform the current state-of-the-art statistical machine translation-based approach, both in terms of grammaticality and fluency.

arXiv:1801.08337

Continuous Space Reordering Models for Phrase-based MT

Bilingual sequence models improve phrase-based translation and reordering by overcoming phrasal independence assumption and handling long range reordering. However, due to data sparsity, these models often fall back to very small context sizes. This problem has been previously addressed by learning sequences over generalized representations such as POS tags or word clusters. In this paper, we explore an alternative based on neural network models. More concretely we train neuralized versions of lexicalized reordering and the operation sequence models using feed-forward neural network. Our results show improvements of up to 0.6 and 0.5 BLEU points on top of the baseline German->English and English->German systems. We also observed improvements compared to the systems that used POS tags and word clusters to train these models. Because we modify the bilingual corpus to integrate reordering operations, this allows us to also train a sequence-to-sequence neural MT model having explicit reordering triggers. Our motivation was to directly enable reordering information in the encoder-decoder framework, which otherwise relies solely on the attention model to handle long range reordering. We tried both coarser and fine-grained reordering operations. However, these experiments did not yield any improvements over the baseline Neural MT systems.

arXiv:1706.00409공식 코드 · Papers with Code

Fader Networks: Manipulating Images by Sliding Attributes

This paper introduces a new encoder-decoder architecture that is trained to reconstruct images by disentangling the salient information of the image and the values of attributes directly in the latent space. As a result, after training, our model can generate different realistic versions of an input image by varying the attribute values. By using continuous attribute values, we can choose how much a specific attribute is perceivable in the generated image. This property could allow for applications where users can modify an image using sliding knobs, like faders on a mixing console, to change the facial expression of a portrait, or to update the color of some objects. Compared to the state-of-the-art which mostly relies on training adversarial networks in pixel space by altering attribute values at train time, our approach results in much simpler training schemes and nicely scales to multiple attributes. We present evidence that our model can significantly change the perceived value of the attributes while preserving the naturalness of images.

arXiv:1801.10198공식 코드 · Papers with Code

Generating Wikipedia by Summarizing Long Sequences

We show that generating English Wikipedia articles can be approached as a multi- document summarization of source documents. We use extractive summarization to coarsely identify salient information and a neural abstractive model to generate the article. For the abstractive model, we introduce a decoder-only architecture that can scalably attend to very long sequences, much longer than typical encoder- decoder architectures used in sequence transduction. We show that this model can generate fluent, coherent multi-sentence paragraphs and even whole Wikipedia articles. When given reference documents, we show it can extract relevant factual information as reflected in perplexity, ROUGE scores and human evaluations.

arXiv:1801.03530

Multi-Scale Attention with Dense Encoder for Handwritten Mathematical Expression Recognition

Handwritten mathematical expression recognition is a challenging problem due to the complicated two-dimensional structures, ambiguous handwriting input and variant scales of handwritten math symbols. To settle this problem, we utilize the attention based encoder-decoder model that recognizes mathematical expression images from two-dimensional layouts to one-dimensional LaTeX strings. We improve the encoder by employing densely connected convolutional networks as they can strengthen feature extraction and facilitate gradient propagation especially on a small training set. We also present a novel multi-scale attention model which is employed to deal with the recognition of math symbols in different scales and save the fine-grained details that will be dropped by pooling operations. Validated on the CROHME competition task, the proposed method significantly outperforms the state-of-the-art methods with an expression recognition accuracy of 52.8% on CROHME 2014 and 50.1% on CROHME 2016, by only using the official training dataset.

arXiv:1711.00727공식 코드 · Papers with Code

Performance Evaluation of Channel Decoding With Deep Neural Networks

With the demand of high data rate and low latency in fifth generation (5G), deep neural network decoder (NND) has become a promising candidate due to its capability of one-shot decoding and parallel computing. In this paper, three types of NND, i.e., multi-layer perceptron (MLP), convolution neural network (CNN) and recurrent neural network (RNN), are proposed with the same parameter magnitude. The performance of these deep neural networks are evaluated through extensive simulation. Numerical results show that RNN has the best decoding performance, yet at the price of the highest computational overhead. Moreover, we find there exists a saturation length for each type of neural network, which is caused by their restricted learning abilities.

arXiv:1802.00094

Single Image Reflection Removal Using Deep Encoder-Decoder Network

Image of a scene captured through a piece of transparent and reflective material, such as glass, is often spoiled by a superimposed layer of reflection image. While separating the reflection from a familiar object in an image is mentally not difficult for humans, it is a challenging, ill-posed problem in computer vision. In this paper, we propose a novel deep convolutional encoder-decoder method to remove the objectionable reflection by learning a map between image pairs with and without reflection. For training the neural network, we model the physical formation of reflections in images and synthesize a large number of photo-realistic reflection-tainted images from reflection-free images collected online. Extensive experimental results show that, although the neural network learns only from synthetic data, the proposed method is effective on real-world images, and it significantly outperforms the other tested state-of-the-art techniques.

arXiv:1801.10517

Densely Dilated Spatial Pooling Convolutional Network using benign loss functions for imbalanced volumetric prostate segmentation

The high incidence rate of prostate disease poses a requirement in early detection for diagnosis. As one of the main imaging methods used for prostate cancer detection, Magnetic Resonance Imaging (MRI) has wide range of appearance and imbalance problems, making automated prostate segmentation fundamental but challenging. Here we propose a novel Densely Dilated Spatial Pooling Convolutional Network (DDSP ConNet) in encoder-decoder structure. It employs dense structure to combine dilated convolution and global pooling, thus supplies coarse segmentation results from encoder and decoder subnet and preserves more contextual information. To obtain richer hierarchical feature maps, residual long connection is furtherly adopted to fuse contexture features. Meanwhile, we adopt DSC loss and Jaccard loss functions to train our DDSP ConNet. We surprisingly found and proved that, in contrast to re-weighted cross entropy, DSC loss and Jaccard loss have a lot of benign properties in theory, including symmetry, continuity and differentiability about the parameters of network. Extensive experiments on the MICCAI PROMISE12 challenge dataset have been done to corroborate the effectiveness of our DDSP ConNet with DSC loss and Jaccard loss. Totally, our method achieves a score of 85.78 in the test dataset, outperforming most of other competitors.

arXiv:1801.04356

Feature Space Transfer for Data Augmentation

The problem of data augmentation in feature space is considered. A new architecture, denoted the FeATure TransfEr Network (FATTEN), is proposed for the modeling of feature trajectories induced by variations of object pose. This architecture exploits a parametrization of the pose manifold in terms of pose and appearance. This leads to a deep encoder/decoder network architecture, where the encoder factors into an appearance and a pose predictor. Unlike previous attempts at trajectory transfer, FATTEN can be efficiently trained end-to-end, with no need to train separate feature transfer functions. This is realized by supplying the decoder with information about a target pose and the use of a multi-task loss that penalizes category- and pose-mismatches. In result, FATTEN discourages discontinuous or non-smooth trajectories that fail to capture the structure of the pose manifold, and generalizes well on object recognition tasks involving large pose variation. Experimental results on the artificial ModelNet database show that it can successfully learn to map source features to target features of a desired pose, while preserving class identity. Most notably, by using feature space transfer for data augmentation (w.r.t. pose and depth) on SUN-RGBD objects, we demonstrate considerable performance improvements on one/few-shot object recognition in a transfer learning setup, compared to current state-of-the-art methods.

arXiv:1801.05122공식 코드 · Papers with Code

Asynchronous Bidirectional Decoding for Neural Machine Translation

The dominant neural machine translation (NMT) models apply unified attentional encoder-decoder neural networks for translation. Traditionally, the NMT decoders adopt recurrent neural networks (RNNs) to perform translation in a left-toright manner, leaving the target-side contexts generated from right to left unexploited during translation. In this paper, we equip the conventional attentional encoder-decoder NMT framework with a backward decoder, in order to explore bidirectional decoding for NMT. Attending to the hidden state sequence produced by the encoder, our backward decoder first learns to generate the target-side hidden state sequence from right to left. Then, the forward decoder performs translation in the forward direction, while in each translation prediction timestep, it simultaneously applies two attention models to consider the source-side and reverse target-side hidden states, respectively. With this new architecture, our model is able to fully exploit source- and target-side contexts to improve translation quality altogether. Experimental results on NIST Chinese-English and WMT English-German translation tasks demonstrate that our model achieves substantial improvements over the conventional NMT by 3.14 and 1.38 BLEU points, respectively. The source code of this work can be obtained from https://github.com/DeepLearnXMU/ABDNMT.

arXiv:1707.05928

Deep Active Learning for Named Entity Recognition

Deep learning has yielded state-of-the-art performance on many natural language processing tasks including named entity recognition (NER). However, this typically requires large amounts of labeled data. In this work, we demonstrate that the amount of labeled training data can be drastically reduced when deep learning is combined with active learning. While active learning is sample-efficient, it can be computationally expensive since it requires iterative retraining. To speed this up, we introduce a lightweight architecture for NER, viz., the CNN-CNN-LSTM model consisting of convolutional character and word encoders and a long short term memory (LSTM) tag decoder. The model achieves nearly state-of-the-art performance on standard datasets for the task while being computationally much more efficient than best performing models. We carry out incremental active learning, during the training process, and are able to nearly match state-of-the-art performance with just 25\% of the original training data.

arXiv:1802.01817

Byte-Level Recursive Convolutional Auto-Encoder for Text

This article proposes to auto-encode text at byte-level using convolutional networks with a recursive architecture. The motivation is to explore whether it is possible to have scalable and homogeneous text generation at byte-level in a non-sequential fashion through the simple task of auto-encoding. We show that non-sequential text generation from a fixed-length representation is not only possible, but also achieved much better auto-encoding results than recurrent networks. The proposed model is a multi-stage deep convolutional encoder-decoder framework using residual connections, containing up to 160 parameterized layers. Each encoder or decoder contains a shared group of modules that consists of either pooling or upsampling layers, making the network recursive in terms of abstraction levels in representation. Results for 6 large-scale paragraph datasets are reported, in 3 languages including Arabic, Chinese and English. Analyses are conducted to study several properties of the proposed model.