PyTorchKR Papers

최신 논문

총 9,642편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:1910.13106

Incorporating Interlocutor-Aware Context into Response Generation on Multi-Party Chatbots

Conventional chatbots focus on two-party response generation, which simplifies the real dialogue scene. In this paper, we strive toward a novel task of Response Generation on Multi-Party Chatbot (RGMPC), where the generated responses heavily rely on the interlocutors' roles (e.g., speaker and addressee) and their utterances. Unfortunately, complex interactions among the interlocutors' roles make it challenging to precisely capture conversational contexts and interlocutors' information. Facing this challenge, we present a response generation model which incorporates Interlocutor-aware Contexts into Recurrent Encoder-Decoder frameworks (ICRED) for RGMPC. Specifically, we employ interactive representations to capture dialogue contexts for different interlocutors. Moreover, we leverage an addressee memory to enhance contextual interlocutor information for the target addressee. Finally, we construct a corpus for RGMPC based on an existing open-access dataset. Automatic and manual evaluations demonstrate that the ICRED remarkably outperforms strong baselines.

arXiv:1910.13062

Disentangling the Spatial Structure and Style in Conditional VAE

This paper aims to disentangle the latent space in cVAE into the spatial structure and the style code, which are complementary to each other, with one of them $z_s$ being label relevant and the other $z_u$ irrelevant. The generator is built by a connected encoder-decoder and a label condition mapping network. Depending on whether the label is related with the spatial structure, the output $z_s$ from the condition mapping network is used either as a style code or a spatial structure code. The encoder provides the label irrelevant posterior from which $z_u$ is sampled. The decoder employs $z_s$ and $z_u$ in each layer by adaptive normalization like SPADE or AdaIN. Extensive experiments on two datasets with different types of labels show the effectiveness of our method.

arXiv:1910.13042

Deep Multi-Magnification Networks for Multi-Class Breast Cancer Image Segmentation

Pathologic analysis of surgical excision specimens for breast carcinoma is important to evaluate the completeness of surgical excision and has implications for future treatment. This analysis is performed manually by pathologists reviewing histologic slides prepared from formalin-fixed tissue. In this paper, we present Deep Multi-Magnification Network trained by partial annotation for automated multi-class tissue segmentation by a set of patches from multiple magnifications in digitized whole slide images. Our proposed architecture with multi-encoder, multi-decoder, and multi-concatenation outperforms other single and multi-magnification-based architectures by achieving the highest mean intersection-over-union, and can be used to facilitate pathologists' assessments of breast cancer.

arXiv:1910.12048

A Deep Learning Approach to Universal Binary Visible Light Communication Transceiver

This paper studies a deep learning (DL) framework for the design of binary modulated visible light communication (VLC) transceiver with universal dimming support. The dimming control for the optical binary signal boils down to a combinatorial codebook design so that the average Hamming weight of binary codewords matches with arbitrary dimming target. An unsupervised DL technique is employed for obtaining a neural network to replace the encoder-decoder pair that recovers the message from the optically transmitted signal. In such a task, a novel stochastic binarization method is developed to generate the set of binary codewords from continuous-valued neural network outputs. For universal support of arbitrary dimming target, the DL-based VLC transceiver is trained with multiple dimming constraints, which turns out to be a constrained training optimization that is very challenging to handle with existing DL methods. We develop a new training algorithm that addresses the dimming constraints through a dual formulation of the optimization. Based on the developed algorithm, the resulting VLC transceiver can be optimized via the end-to-end training procedure. Numerical results verify that the proposed codebook outperforms theoretically best constant weight codebooks under various VLC setups.

arXiv:1910.12010

Dense Dilated Network with Probability Regularized Walk for Vessel Detection

The detection of retinal vessel is of great importance in the diagnosis and treatment of many ocular diseases. Many methods have been proposed for vessel detection. However, most of the algorithms neglect the connectivity of the vessels, which plays an important role in the diagnosis. In this paper, we propose a novel method for retinal vessel detection. The proposed method includes a dense dilated network to get an initial detection of the vessels and a probability regularized walk algorithm to address the fracture issue in the initial detection. The dense dilated network integrates newly proposed dense dilated feature extraction blocks into an encoder-decoder structure to extract and accumulate features at different scales. A multiscale Dice loss function is adopted to train the network. To improve the connectivity of the segmented vessels, we also introduce a probability regularized walk algorithm to connect the broken vessels. The proposed method has been applied on three public data sets: DRIVE, STARE and CHASE_DB1. The results show that the proposed method outperforms the state-of-the-art methods in accuracy, sensitivity, specificity and also are under receiver operating characteristic curve.

arXiv:1910.11910

Learning audio representations via phase prediction

We learn audio representations by solving a novel self-supervised learning task, which consists of predicting the phase of the short-time Fourier transform from its magnitude. A convolutional encoder is used to map the magnitude spectrum of the input waveform to a lower dimensional embedding. A convolutional decoder is then used to predict the instantaneous frequency (i.e., the temporal rate of change of the phase) from such embedding. To evaluate the quality of the learned representations, we evaluate how they transfer to a wide variety of downstream audio tasks. Our experiments reveal that the phase prediction task leads to representations that generalize across different tasks, partially bridging the gap with fully-supervised models. In addition, we show that the predicted phase can be used as initialization of the Griffin-Lim algorithm, thus reducing the number of iterations needed to reconstruct the waveform in the time domain.

arXiv:1910.11871

Towards Online End-to-end Transformer Automatic Speech Recognition

The Transformer self-attention network has recently shown promising performance as an alternative to recurrent neural networks in end-to-end (E2E) automatic speech recognition (ASR) systems. However, Transformer has a drawback in that the entire input sequence is required to compute self-attention. We have proposed a block processing method for the Transformer encoder by introducing a context-aware inheritance mechanism. An additional context embedding vector handed over from the previously processed block helps to encode not only local acoustic information but also global linguistic, channel, and speaker attributes. In this paper, we extend it towards an entire online E2E ASR system by introducing an online decoding process inspired by monotonic chunkwise attention (MoChA) into the Transformer decoder. Our novel MoChA training and inference algorithms exploit the unique properties of Transformer, whose attentions are not always monotonic or peaky, and have multiple heads and residual connections of the decoder layers. Evaluations of the Wall Street Journal (WSJ) and AISHELL-1 show that our proposed online Transformer decoder outperforms conventional chunkwise approaches.

arXiv:1910.12418

Unsupervised pre-training for sequence to sequence speech recognition

This paper proposes a novel approach to pre-train encoder-decoder sequence-to-sequence (seq2seq) model with unpaired speech and transcripts respectively. Our pre-training method is divided into two stages, named acoustic pre-trianing and linguistic pre-training. In the acoustic pre-training stage, we use a large amount of speech to pre-train the encoder by predicting masked speech feature chunks with its context. In the linguistic pre-training stage, we generate synthesized speech from a large number of transcripts using a single-speaker text to speech (TTS) system, and use the synthesized paired data to pre-train decoder. This two-stage pre-training method integrates rich acoustic and linguistic knowledge into seq2seq model, which will benefit downstream automatic speech recognition (ASR) tasks. The unsupervised pre-training is finished on AISHELL-2 dataset and we apply the pre-trained model to multiple paired data ratios of AISHELL-1 and HKUST. We obtain relative character error rate reduction (CERR) from 38.24% to 7.88% on AISHELL-1 and from 12.00% to 1.20% on HKUST. Besides, we apply our pretrained model to a cross-lingual case with CALLHOME dataset. For all six languages in CALLHOME dataset, our pre-training method makes model outperform baseline consistently.

arXiv:1910.12368

Multitask Learning For Different Subword Segmentations In Neural Machine Translation

In Neural Machine Translation (NMT) the usage of subwords and characters as source and target units offers a simple and flexible solution for translation of rare and unseen words. However, selecting the optimal subword segmentation involves a trade-off between expressiveness and flexibility, and is language and dataset-dependent. We present Block Multitask Learning (BMTL), a novel NMT architecture that predicts multiple targets of different granularities simultaneously, removing the need to search for the optimal segmentation strategy. Our multi-task model exhibits improvements of up to 1.7 BLEU points on each decoder over single-task baseline models with the same number of parameters on datasets from two language pairs of IWSLT15 and one from IWSLT19. The multiple hypotheses generated at different granularities can be combined as a post-processing step to give better translations, which improves over hypothesis combination from baseline models while using substantially fewer parameters.

arXiv:1910.12367

Training ASR models by Generation of Contextual Information

Supervised ASR models have reached unprecedented levels of accuracy, thanks in part to ever-increasing amounts of labelled training data. However, in many applications and locales, only moderate amounts of data are available, which has led to a surge in semi- and weakly-supervised learning research. In this paper, we conduct a large-scale study evaluating the effectiveness of weakly-supervised learning for speech recognition by using loosely related contextual information as a surrogate for ground-truth labels. For weakly supervised training, we use 50k hours of public English social media videos along with their respective titles and post text to train an encoder-decoder transformer model. Our best encoder-decoder models achieve an average of 20.8% WER reduction over a 1000 hours supervised baseline, and an average of 13.4% WER reduction when using only the weakly supervised encoder for CTC fine-tuning. Our results show that our setup for weak supervision improved both the encoder acoustic representations as well as the decoder language generation abilities.

arXiv:1910.12279공식 코드 · Papers with Code

Memeify: A Large-Scale Meme Generation System

Interest in the research areas related to meme propagation and generation has been increasing rapidly in the last couple of years. Meme datasets available online are either specific to a context or contain no class information. Here, we prepare a large-scale dataset of memes with captions and class labels. The dataset consists of 1.1 million meme captions from 128 classes. We also provide reasoning for the existence of broad categories, called "themes" across the meme dataset; each theme consists of multiple meme classes. Our generation system uses a trained state-of-the-art transformer-based model for caption generation by employing an encoder-decoder architecture. We develop a web interface, called Memeify for users to generate memes of their choice, and explain in detail, the working of individual components of the system. We also perform a qualitative evaluation of the generated memes by conducting a user study. A link to the demonstration of the Memeify system is https://youtu.be/P_Tfs0X-czs.

arXiv:1910.12740공식 코드 · Papers with Code

Sequence-to-sequence Automatic Speech Recognition with Word Embedding Regularization and Fused Decoding

In this paper, we investigate the benefit that off-the-shelf word embedding can bring to the sequence-to-sequence (seq-to-seq) automatic speech recognition (ASR). We first introduced the word embedding regularization by maximizing the cosine similarity between a transformed decoder feature and the target word embedding. Based on the regularized decoder, we further proposed the fused decoding mechanism. This allows the decoder to consider the semantic consistency during decoding by absorbing the information carried by the transformed decoder feature, which is learned to be close to the target word embedding. Initial results on LibriSpeech demonstrated that pre-trained word embedding can significantly lower ASR recognition error with a negligible cost, and the choice of word embedding algorithms among Skip-gram, CBOW and BERT is important.

arXiv:1910.12699

Self-supervised learning of class embeddings from video

This work explores how to use self-supervised learning on videos to learn a class-specific image embedding that encodes pose and shape information. At train time, two frames of the same video of an object class (e.g. human upper body) are extracted and each encoded to an embedding. Conditioned on these embeddings, the decoder network is tasked to transform one frame into another. To successfully perform long range transformations (e.g. a wrist lowered in one image should be mapped to the same wrist raised in another), we introduce a hierarchical probabilistic network decoder model. Once trained, the embedding can be used for a variety of downstream tasks and domains. We demonstrate our approach quantitatively on three distinct deformable object classes -- human full bodies, upper bodies, faces -- and show experimentally that the learned embeddings do indeed generalise. They achieve state-of-the-art performance in comparison to other self-supervised methods trained on the same datasets, and approach the performance of fully supervised methods.

arXiv:1910.11483

Generating a Common Question from Multiple Documents using Multi-source Encoder-Decoder Models

Ambiguous user queries in search engines result in the retrieval of documents that often span multiple topics. One potential solution is for the search engine to generate multiple refined queries, each of which relates to a subset of the documents spanning the same topic. A preliminary step towards this goal is to generate a question that captures common concepts of multiple documents. We propose a new task of generating common question from multiple documents and present simple variant of an existing multi-source encoder-decoder framework, called the Multi-Source Question Generator (MSQG). We first train an RNN-based single encoder-decoder generator from (single document, question) pairs. At test time, given multiple documents, the 'Distribute' step of our MSQG model predicts target word distributions for each document using the trained model. The 'Aggregate' step aggregates these distributions to generate a common question. This simple yet effective strategy significantly outperforms several existing baseline models applied to the new task when evaluated using automated metrics and human judgments on the MS-MARCO-QA dataset.

arXiv:1910.11456

Unified Multi-scale Feature Abstraction for Medical Image Segmentation

Automatic medical image segmentation, an essential component of medical image analysis, plays an importantrole in computer-aided diagnosis. For example, locating and segmenting the liver can be very helpful in livercancer diagnosis and treatment. The state-of-the-art models in medical image segmentation are variants ofthe encoder-decoder architecture such as fully convolutional network (FCN) and U-Net.1A major focus ofthe FCN based segmentation methods has been on network structure engineering by incorporating the latestCNN structures such as ResNet2and DenseNet.3In addition to exploring new network structures for efficientlyabstracting high level features, incorporating structures for multi-scale image feature extraction in FCN hashelped to improve performance in segmentation tasks. In this paper, we design a new multi-scale networkarchitecture, which takes multi-scale inputs with dedicated convolutional paths to efficiently combine featuresfrom different scales to better utilize the hierarchical information.

arXiv:1910.11764공식 코드 · Papers with Code

ClsGAN: Selective Attribute Editing Model Based On Classification Adversarial Network

Attribution editing has achieved remarkable progress in recent years owing to the encoder-decoder structure and generative adversarial network (GAN). However, it remains challenging in generating high-quality images with accurate attribute transformation. Attacking these problems, the work proposes a novel selective attribute editing model based on classification adversarial network (referred to as ClsGAN) that shows good balance between attribute transfer accuracy and photo-realistic images. Considering that the editing images are prone to be affected by original attribute due to skip-connection in encoder-decoder structure, an upper convolution residual network (referred to as Tr-resnet) is presented to selectively extract information from the source image and target label. In addition, to further improve the transfer accuracy of generated images, an attribute adversarial classifier (referred to as Atta-cls) is introduced to guide the generator from the perspective of attribute through learning the defects of attribute transfer images. Experimental results on CelebA demonstrate that our ClsGAN performs favorably against state-of-the-art approaches in image quality and transfer accuracy. Moreover, ablation studies are also designed to verify the great performance of Tr-resnet and Atta-cls.

arXiv:1910.11067공식 코드 · Papers with Code

Supervised Encoding for Discrete Representation Learning

Classical supervised classification tasks search for a nonlinear mapping that maps each encoded feature directly to a probability mass over the labels. Such a learning framework typically lacks the intuition that encoded features from the same class tend to be similar and thus has little interpretability for the learned features. In this paper, we propose a novel supervised learning model named Supervised-Encoding Quantizer (SEQ). The SEQ applies a quantizer to cluster and classify the encoded features. We found that the quantizer provides an interpretable graph where each cluster in the graph represents a class of data samples that have a particular style. We also trained a decoder that can decode convex combinations of the encoded features from similar and different clusters and provide guidance on style transfer between sub-classes.

arXiv:1910.11030공식 코드 · Papers with Code

Spatiotemporal Tile-based Attention-guided LSTMs for Traffic Video Prediction

This extended abstract describes our solution for the Traffic4Cast Challenge 2019. The key problem we addressed is to properly model both low-level (pixel based) and high-level spatial information while still preserve the temporal relations among the frames. Our approach is inspired by the recent adoption of convolutional features into a recurrent neural networks such as LSTM to jointly capture the spatio-temporal dependency. While this approach has been proven to surpass the traditional stacked CNNs (using 2D or 3D kernels) in action recognition, we observe suboptimal performance in traffic prediction setting. Therefore, we apply a number of adaptations in the frame encoder-decoder layers and in sampling procedure to better capture the high-resolution trajectories, and to increase the training efficiency.

arXiv:1910.11250

When Segmentation is Not Enough: Rectifying Visual-Volume Discordance Through Multisensor Depth-Refined Semantic Segmentation for Food Intake Tracking in Long-Term Care

Malnutrition is a multidomain problem affecting 54% of older adults in long-term care (LTC). Monitoring nutritional intake in LTC is laborious and subjective, limiting clinical inference capabilities. Recent advances in automatic image-based food estimation have not yet been evaluated in LTC settings. Here, we describe a fully automatic imaging system for quantifying food intake. We propose a novel deep convolutional encoder-decoder food network with depth-refinement (EDFN-D) using an RGB-D camera for quantifying a plate's remaining food volume relative to reference portions in whole and modified texture foods. We trained and validated the network on the pre-labelled UNIMIB2016 food dataset and tested on our two novel LTC-inspired plate datasets (689 plate images, 36 unique foods). EDFN-D performed comparably to depth-refined graph cut on IOU (0.879 vs. 0.887), with intake errors well below typical 50% (mean percent intake error: -4.2%). We identify how standard segmentation metrics are insufficient due to visual-volume discordance, and include volume disparity analysis to facilitate system trust. This system provides improved transparency, approximates human assessors with enhanced objectivity, accuracy, and precision while avoiding hefty semi-automatic method time requirements. This may help address short-comings currently limiting utility of automated early malnutrition detection in resource-constrained LTC and hospital settings.

arXiv:1910.11161

Diversifying Topic-Coherent Response Generation for Natural Multi-turn Conversations

Although response generation (RG) diversification for single-turn dialogs has been well developed, it is less investigated for natural multi-turn conversations. Besides, past work focused on diversifying responses without considering topic coherence to the context, producing uninformative replies. In this paper, we propose the Topic-coherent Hierarchical Recurrent Encoder-Decoder model (THRED) to diversify the generated responses without deviating the contextual topics for multi-turn conversations. In overall, we build a sequence-to-sequence net (Seq2Seq) to model multi-turn conversations. And then we resort to the latent Variable Hierarchical Recurrent Encoder-Decoder model (VHRED) to learn global contextual distribution of dialogs. Besides, we construct a dense topic matrix which implies word-level correlations of the conversation corpora. The topic matrix is used to learn local topic distribution of the contextual utterances. By incorporating both the global contextual distribution and the local topic distribution, THRED produces both diversified and topic-coherent replies. In addition, we propose an explicit metric (\emph{TopicDiv}) to measure the topic divergence between the post and generated response, and we also propose an overall metric combining the diversification metric (\emph{Distinct}) and \emph{TopicDiv}. We evaluate our model comparing with three baselines (Seq2Seq, HRED and VHRED) on two real-world corpora, respectively, and demonstrate its outstanding performance in both diversification and topic coherence.