PyTorchKR Papers

최신 논문

총 9,642편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2010.12829

Multilingual Speech Translation with Efficient Finetuning of Pretrained Models

We present a simple yet effective approach to build multilingual speech-to-text (ST) translation by efficient transfer learning from pretrained speech encoder and text decoder. Our key finding is that a minimalistic LNA (LayerNorm and Attention) finetuning can achieve zero-shot crosslingual and cross-modality transfer ability by only finetuning less than 10% of the pretrained parameters. This enables effectively leveraging large pretrained models with low training cost. Using wav2vec 2.0 for acoustic modeling, and mBART for multilingual text generation, our approach advanced the new state-of-the-art for 34 translation directions (and surpassing cascaded ST for 23 of them) on large-scale multilingual ST benchmark CoVoST 2 (+6.4 BLEU on average across 15 En-X directions and +5.1 BLEU on average across 19 X-En directions). Our approach demonstrates strong zero-shot performance in a many-to-many multilingual model (+5.7 BLEU on average across 18 non-English directions), making it an appealing approach for attaining high-quality speech translation with improved parameter and data efficiency.

arXiv:2010.12828

Keyphrase Extraction with Dynamic Graph Convolutional Networks and Diversified Inference

Keyphrase extraction (KE) aims to summarize a set of phrases that accurately express a concept or a topic covered in a given document. Recently, Sequence-to-Sequence (Seq2Seq) based generative framework is widely used in KE task, and it has obtained competitive performance on various benchmarks. The main challenges of Seq2Seq methods lie in acquiring informative latent document representation and better modeling the compositionality of the target keyphrases set, which will directly affect the quality of generated keyphrases. In this paper, we propose to adopt the Dynamic Graph Convolutional Networks (DGCN) to solve the above two problems simultaneously. Concretely, we explore to integrate dependency trees with GCN for latent representation learning. Moreover, the graph structure in our model is dynamically modified during the learning process according to the generated keyphrases. To this end, our approach is able to explicitly learn the relations within the keyphrases collection and guarantee the information interchange between encoder and decoder in both directions. Extensive experiments on various KE benchmark datasets demonstrate the effectiveness of our approach.

arXiv:2010.12827

Context-aware Decoder for Neural Machine Translation using a Target-side Document-Level Language Model

Although many context-aware neural machine translation models have been proposed to incorporate contexts in translation, most of those models are trained end-to-end on parallel documents aligned in sentence-level. Because only a few domains (and language pairs) have such document-level parallel data, we cannot perform accurate context-aware translation in most domains. We therefore present a simple method to turn a sentence-level translation model into a context-aware model by incorporating a document-level language model into the decoder. Our context-aware decoder is built upon only a sentence-level parallel corpora and monolingual corpora; thus no document-level parallel data is needed. In a theoretical viewpoint, the core part of this work is the novel representation of contextual information using point-wise mutual information between context and the current sentence. We show the effectiveness of our approach in three language pairs, English to French, English to Russian, and Japanese to English, by evaluation in \textsc{bleu} and contrastive tests for context-aware translation.

arXiv:2010.12780

Open-Domain Dialogue Generation Based on Pre-trained Language Models

Pre-trained language models have been successfully used in response generation for open-domain dialogue. Four main frameworks have been proposed: (1) Transformer-ED using Transformer encoder and decoder separately for source and target sentences; (2) Transformer-Dec using Transformer decoder for both source and target sentences; (3) Transformer-MLM using Transformer decoder that applies bi-directional attention on the source side and left-to-right attention on the target side with masked language model objective; and (4) Transformer-AR that uses auto-regressive objective instead. In this study, we compare these frameworks on 3 datasets, and our comparison reveals that the best framework uses bidirectional attention on the source side and does not separate encoder and decoder. We also examine model discrepancy, and our experiments confirm that the performance of a model is directly impacted by the underlying discrepancies. We then propose two correction methods to reduce the discrepancies, and both improve the model performance. These results show that discrepancies is an important factor to consider when we use a pre-trained model, and a reduction in discrepancies can lead to improved performance.

arXiv:2010.12770공식 코드 · Papers with Code

Conversational Semantic Parsing for Dialog State Tracking

We consider a new perspective on dialog state tracking (DST), the task of estimating a user's goal through the course of a dialog. By formulating DST as a semantic parsing task over hierarchical representations, we can incorporate semantic compositionality, cross-domain knowledge sharing and co-reference. We present TreeDST, a dataset of 27k conversations annotated with tree-structured dialog states and system acts. We describe an encoder-decoder framework for DST with hierarchical representations, which leads to 20% improvement over state-of-the-art DST approaches that operate on a flat meaning space of slot-value pairs.

arXiv:2010.13106공식 코드 · Papers with Code

Scribble-based Weakly Supervised Deep Learning for Road Surface Extraction from Remote Sensing Images

Road surface extraction from remote sensing images using deep learning methods has achieved good performance, while most of the existing methods are based on fully supervised learning, which requires a large amount of training data with laborious per-pixel annotation. In this paper, we propose a scribble-based weakly supervised road surface extraction method named ScRoadExtractor, which learns from easily accessible scribbles such as centerlines instead of densely annotated road surface ground-truths. To propagate semantic information from sparse scribbles to unlabeled pixels, we introduce a road label propagation algorithm which considers both the buffer-based properties of road networks and the color and spatial information of super-pixels. The proposal masks generated from the road label propagation algorithm are utilized to train a dual-branch encoder-decoder network we designed, which consists of a semantic segmentation branch and an auxiliary boundary detection branch. We perform experiments on three diverse road datasets that are comprised of highresolution remote sensing satellite and aerial images across the world. The results demonstrate that ScRoadExtractor exceed the classic scribble-supervised segmentation method by 20% for the intersection over union (IoU) indicator and outperform the state-of-the-art scribble-based weakly supervised methods at least 4%.

arXiv:2010.13082

Context Aware 3D UNet for Brain Tumor Segmentation

Deep convolutional neural network (CNN) achieves remarkable performance for medical image analysis. UNet is the primary source in the performance of 3D CNN architectures for medical imaging tasks, including brain tumor segmentation. The skip connection in the UNet architecture concatenates features from both encoder and decoder paths to extract multi-contextual information from image data. The multi-scaled features play an essential role in brain tumor segmentation. However, the limited use of features can degrade the performance of the UNet approach for segmentation. In this paper, we propose a modified UNet architecture for brain tumor segmentation. In the proposed architecture, we used densely connected blocks in both encoder and decoder paths to extract multi-contextual information from the concept of feature reusability. In addition, residual-inception blocks (RIB) are used to extract the local and global information by merging features of different kernel sizes. We validate the proposed architecture on the multi-modal brain tumor segmentation challenge (BRATS) 2020 testing dataset. The dice (DSC) scores of the whole tumor (WT), tumor core (TC), and enhancing tumor (ET) are 89.12%, 84.74%, and 79.12%, respectively.

arXiv:2010.13047

Orthros: Non-autoregressive End-to-end Speech Translation with Dual-decoder

Fast inference speed is an important goal towards real-world deployment of speech translation (ST) systems. End-to-end (E2E) models based on the encoder-decoder architecture are more suitable for this goal than traditional cascaded systems, but their effectiveness regarding decoding speed has not been explored so far. Inspired by recent progress in non-autoregressive (NAR) methods in text-based translation, which generates target tokens in parallel by eliminating conditional dependencies, we study the problem of NAR decoding for E2E-ST. We propose a novel NAR E2E-ST framework, Orthros, in which both NAR and autoregressive (AR) decoders are jointly trained on the shared speech encoder. The latter is used for selecting better translation among various length candidates generated from the former, which dramatically improves the effectiveness of a large length beam with negligible overhead. We further investigate effective length prediction methods from speech inputs and the impact of vocabulary sizes. Experiments on four benchmarks show the effectiveness of the proposed method in improving inference speed while maintaining competitive translation quality compared to state-of-the-art AR E2E-ST systems.

arXiv:2010.12973

Unsupervised Learning of Disentangled Speech Content and Style Representation

We present an approach for unsupervised learning of speech representation disentangling contents and styles. Our model consists of: (1) a local encoder that captures per-frame information; (2) a global encoder that captures per-utterance information; and (3) a conditional decoder that reconstructs speech given local and global latent variables. Our experiments show that (1) the local latent variables encode speech contents, as reconstructed speech can be recognized by ASR with low word error rates (WER), even with a different global encoding; (2) the global latent variables encode speaker style, as reconstructed speech shares speaker identity with the source utterance of the global encoding. Additionally, we demonstrate an useful application from our pre-trained model, where we can train a speaker recognition model from the global latent variables and achieve high accuracy by fine-tuning with as few data as one label per speaker.

arXiv:2010.13418

Residual Recurrent CRNN for End-to-End Optical Music Recognition on Monophonic Scores

One of the challenges of the Optical Music Recognition task is to transcript the symbols of the camera-captured images into digital music notations. Previous end-to-end model which was developed as a Convolutional Recurrent Neural Network does not explore sufficient contextual information from full scales and there is still a large room for improvement. We propose an innovative framework that combines a block of Residual Recurrent Convolutional Neural Network with a recurrent Encoder-Decoder network to map a sequence of monophonic music symbols corresponding to the notations present in the image. The Residual Recurrent Convolutional block can improve the ability of the model to enrich the context information. The experiment results are benchmarked against a publicly available dataset called CAMERA-PRIMUS, which demonstrates that our approach surpass the state-of-the-art end-to-end method using Convolutional Recurrent Neural Network.

arXiv:2010.13982공식 코드 · Papers with Code

Predict and Use Latent Patterns for Short-Text Conversation

Many neural network models nowadays have achieved promising performances in Chit-chat settings. The majority of them rely on an encoder for understanding the post and a decoder for generating the response. Without given assigned semantics, the models lack the fine-grained control over responses as the semantic mapping between posts and responses is hidden on the fly within the end-to-end manners. Some previous works utilize sampled latent words as a controllable semantic form to drive the generated response around the work, but few works attempt to use more complex semantic patterns to guide the generation. In this paper, we propose to use more detailed semantic forms, including latent responses and part-of-speech sequences sampled from the corresponding distributions, as the controllable semantics to guide the generation. Our results show that the richer semantics are not only able to provide informative and diverse responses, but also increase the overall performance of response quality, including fluency and coherence.

arXiv:2010.13903

T$^2$-Net: A Semi-supervised Deep Model for Turbulence Forecasting

Accurate air turbulence forecasting can help airlines avoid hazardous turbulence, guide the routes that keep passengers safe, maximize efficiency, and reduce costs. Traditional turbulence forecasting approaches heavily rely on painstakingly customized turbulence indexes, which are less effective in dynamic and complex weather conditions. The recent availability of high-resolution weather data and turbulence records allows more accurate forecasting of the turbulence in a data-driven way. However, it is a non-trivial task for developing a machine learning based turbulence forecasting system due to two challenges: (1) Complex spatio-temporal correlations, turbulence is caused by air movement with complex spatio-temporal patterns, (2) Label scarcity, very limited turbulence labels can be obtained. To this end, in this paper, we develop a unified semi-supervised framework, T$^2$-Net, to address the above challenges. Specifically, we first build an encoder-decoder paradigm based on the convolutional LSTM to model the spatio-temporal correlations. Then, to tackle the label scarcity problem, we propose a novel Dual Label Guessing method to take advantage of massive unlabeled turbulence data. It integrates complementary signals from the main Turbulence Forecasting task and the auxiliary Turbulence Detection task to generate pseudo-labels, which are dynamically utilized as additional training data. Finally, extensive experimental results on a real-world turbulence dataset validate the superiority of our method on turbulence forecasting.

arXiv:2010.14481공식 코드 · Papers with Code

Fast Interleaved Bidirectional Sequence Generation

Independence assumptions during sequence generation can speed up inference, but parallel generation of highly inter-dependent tokens comes at a cost in quality. Instead of assuming independence between neighbouring tokens (semi-autoregressive decoding, SA), we take inspiration from bidirectional sequence generation and introduce a decoder that generates target words from the left-to-right and right-to-left directions simultaneously. We show that we can easily convert a standard architecture for unidirectional decoding into a bidirectional decoder by simply interleaving the two directions and adapting the word positions and self-attention masks. Our interleaved bidirectional decoder (IBDecoder) retains the model simplicity and training efficiency of the standard Transformer, and on five machine translation tasks and two document summarization tasks, achieves a decoding speedup of ~2X compared to autoregressive decoding with comparable quality. Notably, it outperforms left-to-right SA because the independence assumptions in IBDecoder are more felicitous. To achieve even higher speedups, we explore hybrid models where we either simultaneously predict multiple neighbouring tokens per direction, or perform multi-directional decoding by partitioning the target sequence. These methods achieve speedups to 4X-11X across different tasks at the cost of <1 BLEU or <0.5 ROUGE (on average). Source code is released at https://github.com/bzhangGo/zero.

arXiv:2010.14318

Multitask Training with Text Data for End-to-End Speech Recognition

We propose a multitask training method for attention-based end-to-end speech recognition models. We regularize the decoder in a listen, attend, and spell model by multitask training it on both audio-text and text-only data. Trained on the 100-hour subset of LibriSpeech, the proposed method, without requiring an additional language model, leads to an 11% relative performance improvement over the baseline and approaches the performance of language model shallow fusion on the test-clean evaluation set. We observe a similar trend on the whole 960-hour LibriSpeech training set. Analyses of different types of errors and sample output sentences demonstrate that the proposed method can incorporate language level information, suggesting its effectiveness in real-world applications.

arXiv:2010.12042

SAINT+: Integrating Temporal Features for EdNet Correctness Prediction

We propose SAINT+, a successor of SAINT which is a Transformer based knowledge tracing model that separately processes exercise information and student response information. Following the architecture of SAINT, SAINT+ has an encoder-decoder structure where the encoder applies self-attention layers to a stream of exercise embeddings, and the decoder alternately applies self-attention layers and encoder-decoder attention layers to streams of response embeddings and encoder output. Moreover, SAINT+ incorporates two temporal feature embeddings into the response embeddings: elapsed time, the time taken for a student to answer, and lag time, the time interval between adjacent learning activities. We empirically evaluate the effectiveness of SAINT+ on EdNet, the largest publicly available benchmark dataset in the education domain. Experimental results show that SAINT+ achieves state-of-the-art performance in knowledge tracing with an improvement of 1.25% in area under receiver operating characteristic curve compared to SAINT, the current state-of-the-art model in EdNet dataset.

arXiv:2010.12267공식 코드 · Papers with Code

Show and Speak: Directly Synthesize Spoken Description of Images

This paper proposes a new model, referred to as the show and speak (SAS) model that, for the first time, is able to directly synthesize spoken descriptions of images, bypassing the need for any text or phonemes. The basic structure of SAS is an encoder-decoder architecture that takes an image as input and predicts the spectrogram of speech that describes this image. The final speech audio is obtained from the predicted spectrogram via WaveNet. Extensive experiments on the public benchmark database Flickr8k demonstrate that the proposed SAS is able to synthesize natural spoken descriptions for images, indicating that synthesizing spoken descriptions for images while bypassing text and phonemes is feasible.

arXiv:2010.12249공식 코드 · Papers with Code

Multi Scale Identity-Preserving Image-to-Image Translation Network for Low-Resolution Face Recognition

State-of-the-art deep neural network models have reached near perfect face recognition accuracy rates on controlled high-resolution face images. However, their performance is drastically degraded when they are tested with very low-resolution face images. This is particularly critical in surveillance systems, where a low-resolution probe image is to be matched with high-resolution gallery images. super-resolution techniques aim at producing high-resolution face images from low-resolution counterparts. While they are capable of reconstructing images that are visually appealing, the identity-related information is not preserved. Here, we propose an identity-preserving end-to-end image-to-image translation deep neural network which is capable of super-resolving very low-resolution faces to their high-resolution counterparts while preserving identity-related information. We achieved this by training a very deep convolutional encoder-decoder network with a symmetric contracting path between corresponding layers. This network was trained with a combination of a reconstruction and an identity-preserving loss, on multi-scale low-resolution conditions. Extensive quantitative evaluations of our proposed model demonstrated that it outperforms competing super-resolution and low-resolution face recognition methods on natural and artificial low-resolution face data sets and even unseen identities.

arXiv:2010.12219

A Teacher-Student Framework for Semi-supervised Medical Image Segmentation From Mixed Supervision

Standard segmentation of medical images based on full-supervised convolutional networks demands accurate dense annotations. Such learning framework is built on laborious manual annotation with restrict demands for expertise, leading to insufficient high-quality labels. To overcome such limitation and exploit massive weakly labeled data, we relaxed the rigid labeling requirement and developed a semi-supervised learning framework based on a teacher-student fashion for organ and lesion segmentation with partial dense-labeled supervision and supplementary loose bounding-box supervision which are easier to acquire. Observing the geometrical relation of an organ and its inner lesions in most cases, we propose a hierarchical organ-to-lesion (O2L) attention module in a teacher segmentor to produce pseudo-labels. Then a student segmentor is trained with combinations of manual-labeled and pseudo-labeled annotations. We further proposed a localization branch realized via an aggregation of high-level features in a deep decoder to predict locations of organ and lesion, which enriches student segmentor with precise localization information. We validated each design in our model on LiTS challenge datasets by ablation study and showed its state-of-the-art performance compared with recent methods. We show our model is robust to the quality of bounding box and achieves comparable performance compared with full-supervised learning methods.

arXiv:2010.12188

Generating Long Financial Report using Conditional Variational Autoencoders with Knowledge Distillation

Automatically generating financial report from a piece of news is quite a challenging task. Apparently, the difficulty of this task lies in the lack of sufficient background knowledge to effectively generate long financial report. To address this issue, this paper proposes the conditional variational autoencoders (CVAE) based approach which distills external knowledge from a corpus of news-report data. Particularly, we choose Bi-GRU as the encoder and decoder component of CVAE, and learn the latent variable distribution from input news. A higher level latent variable distribution is learnt from a corpus set of news-report data, respectively extr acted for each input news, to provide background knowledge to previously learnt latent variable distribution. Then, a teacher-student network is employed to distill knowledge to refine theoutput of the decoder component. To evaluate the model performance of the proposed approach, extensive experiments are preformed on a public dataset and two widely adopted evaluation criteria, i.e., BLEU and ROUGE, are chosen in the experiment. The promising experimental results demonstrate that the proposed approach is superior to the rest compared methods.

arXiv:2010.12104공식 코드 · Papers with Code

How Phonotactics Affect Multilingual and Zero-shot ASR Performance

The idea of combining multiple languages' recordings to train a single automatic speech recognition (ASR) model brings the promise of the emergence of universal speech representation. Recently, a Transformer encoder-decoder model has been shown to leverage multilingual data well in IPA transcriptions of languages presented during training. However, the representations it learned were not successful in zero-shot transfer to unseen languages. Because that model lacks an explicit factorization of the acoustic model (AM) and language model (LM), it is unclear to what degree the performance suffered from differences in pronunciation or the mismatch in phonotactics. To gain more insight into the factors limiting zero-shot ASR transfer, we replace the encoder-decoder with a hybrid ASR system consisting of a separate AM and LM. Then, we perform an extensive evaluation of monolingual, multilingual, and crosslingual (zero-shot) acoustic and language models on a set of 13 phonetically diverse languages. We show that the gain from modeling crosslingual phonotactics is limited, and imposing a too strong model can hurt the zero-shot transfer. Furthermore, we find that a multilingual LM hurts a multilingual ASR system's performance, and retaining only the target language's phonotactic data in LM training is preferable.