Delineation of the cardiac structures from 2D echocardiographic images is a common clinical task to establish a diagnosis. Over the past decades, the automation of this task has been the subject of intense research. In this paper, we evaluate how far the state-of-the-art encoder-decoder deep convolutional neural network methods can go at assessing 2D echocardiographic images, i.e segmenting cardiac structures as well as estimating clinical indices, on a dataset especially designed to answer this objective. We therefore introduce the Cardiac Acquisitions for Multi-structure Ultrasound Segmentation (CAMUS) dataset, the largest publicly-available and fully-annotated dataset for the purpose of echocardiographic assessment. The dataset contains two and four-chamber acquisitions from 500 patients with reference measurements from one cardiologist on the full dataset and from three cardiologists on a fold of 50 patients. Results show that encoder-decoder based architectures outperform state-of-the-art non-deep learning methods and faithfully reproduce the expert analysis for the end-diastolic and end-systolic left ventricular volumes, with a mean correlation of 0.95 and an absolute mean error of 9.5 ml. Concerning the ejection fraction of the left ventricle, results are more contrasted with a mean correlation coefficient of 0.80 and an absolute mean error of 5.6 %. Although these results are below the inter-observer scores, they remain slightly worse than the intra-observer's ones. Based on this observation, areas for improvement are defined, which open the door for accurate and fully-automatic analysis of 2D echocardiographic images.
The primary goal of skeletal motion prediction is to generate future motion by observing a sequence of 3D skeletons. A key challenge in motion prediction is the fact that a motion can often be performed in several different ways, with each consisting of its own configuration of poses and their spatio-temporal dependencies, and as a result, the predicted poses often converge to the motionless poses or non-human like motions in long-term prediction. This leads us to define a hierarchical recurrent network model that explicitly characterizes these internal configurations of poses and their local and global spatio-temporal dependencies. The model introduces a latent vector variable from the Lie algebra to represent spatial and temporal relations simultaneously. Furthermore, a structured stack LSTM-based decoder is devised to decode the predicted poses with a new loss function defined to estimate the quantized weight of each body part in a pose. Empirical evaluations on benchmark datasets suggest our approach significantly outperforms the state-of-the-art methods on both short-term and long-term motion prediction.
Automatic question generation aims at the generation of questions from a context, with the corresponding answers being sub-spans of the given passage. Whereas, most of the methods mostly rely on heuristic rules to generate questions, more recently also neural network approaches have been proposed. In this work, we propose a variant of the self-attention Transformer network architectures model to generate meaningful and diverse questions. To this end, we propose an easy to use model consisting of the conjunction of the Transformer decoder GPT-2 model with Transformer encoder BERT for the downstream task for question answering. The model is trained in an end-to-end fashion, where the language model is trained to produce a question-answer-aware input representation that facilitates to generate an answer focused question. Our result of neural question generation from text on the SQuAD 1.1 dataset suggests that our method can produce semantically correct and diverse questions. Additionally, we assessed the performance of our proposed method for the downstream task of question answering. The analysis shows that our proposed generation & answering collaboration framework relatively improves both tasks and is particularly powerful in the semi-supervised setup. The results further suggest a robust and comparably lean pipeline facilitating question generation in the small-data regime.
Speech signal is constituted and contributed by various informative factors, such as linguistic content and speaker characteristic. There have been notable recent studies attempting to factorize speech signal into these individual factors without requiring any annotation. These studies typically assume continuous representation for linguistic content, which is not in accordance with general linguistic knowledge and may make the extraction of speaker information less successful. This paper proposes the mixture factorized auto-encoder (mFAE) for unsupervised deep factorization. The encoder part of mFAE comprises a frame tokenizer and an utterance embedder. The frame tokenizer models linguistic content of input speech with a discrete categorical distribution. It performs frame clustering by assigning each frame a soft mixture label. The utterance embedder generates an utterance-level vector representation. A frame decoder serves to reconstruct speech features from the encoders'outputs. The mFAE is evaluated on speaker verification (SV) task and unsupervised subword modeling (USM) task. The SV experiments on VoxCeleb 1 show that the utterance embedder is capable of extracting speaker-discriminative embeddings with performance comparable to a x-vector baseline. The USM experiments on ZeroSpeech 2017 dataset verify that the frame tokenizer is able to capture linguistic content and the utterance embedder can acquire speaker-related information.
With the rapid growth of deep learning in many fields, machine learning-assisted communication systems had attracted lots of researches with many eye-catching initial results. At the present stage, most of the methods still have great demand of massive labeled data for supervised learning. However, obtaining labeled data in the practical applications is not feasible, which may result in severe performance degradation due to channel variations. To overcome such a constraint, syndrome loss has been proposed to penalize non-valid decoded codewords and achieve unsupervised learning for neural network-based decoder. However, it cannot be applied to polar decoder directly. In this work, by exploiting the nature of polar codes, we propose a modified syndrome loss. From simulation results, the proposed method demonstrates that domain-specific knowledge and know-how in code structure can enable unsupervised learning for neural network-based polar decoder.
Generally, Neural Machine Translation models generate target words in a left-to-right (L2R) manner and fail to exploit any future (right) semantics information, which usually produces an unbalanced translation. Recent works attempt to utilize the right-to-left (R2L) decoder in bidirectional decoding to alleviate this problem. In this paper, we propose a novel \textbf{D}ynamic \textbf{I}nteraction \textbf{M}odule (\textbf{DIM}) to dynamically exploit target semantics from R2L translation for enhancing the L2R translation quality. Different from other bidirectional decoding approaches, DIM firstly extracts helpful target information through addressing and reading operations, then updates target semantics for tracking the interactive history. Additionally, we further introduce an \textbf{agreement regularization} term into the training objective to narrow the gap between L2R and R2L translations. Experimental results on NIST Chinese$\Rightarrow$English and WMT'16 English$\Rightarrow$Romanian translation tasks show that our system achieves significant improvements over baseline systems, which also reaches comparable results compared to the state-of-the-art Transformer model with much fewer parameters of it.
Texts from scene images typically consist of several characters and exhibit a characteristic sequence structure. Existing methods capture the structure with the sequence-to-sequence models by an encoder to have the visual representations and then a decoder to translate the features into the label sequence. In this paper, we study text recognition framework by considering the long-term temporal dependencies in the encoder stage. We demonstrate that the proposed Temporal Convolutional Encoder with increased sequential extents improves the accuracy of text recognition. We also study the impact of different attention modules in convolutional blocks for learning accurate text representations. We conduct comparisons on seven datasets and the experiments demonstrate the effectiveness of our proposed approach.
In this paper, we present a multi-class eye segmentation method that can run the hardware limitations for real-time inference. Our approach includes three major stages: get a grayscale image from the input, segment three distinct eye region with a deep network, and remove incorrect areas with heuristic filters. Our model based on the encoder decoder structure with the key is the depthwise convolution operation to reduce the computation cost. We experiment on OpenEDS, a large scale dataset of eye images captured by a head-mounted display with two synchronized eye facing cameras. We achieved the mean intersection over union (mIoU) of 94.85% with a model of size 0.4 megabytes. The source code are available https://github.com/th2l/Eye_VR_Segmentation
We present a novel end-to-end single-shot method that segments countable object instances (things) as well as background regions (stuff) into a non-overlapping panoptic segmentation at almost video frame rate. Current state-of-the-art methods are far from reaching video frame rate and mostly rely on merging instance segmentation with semantic background segmentation, making them impractical to use in many applications such as robotics. Our approach relaxes this requirement by using an object detector but is still able to resolve inter- and intra-class overlaps to achieve a non-overlapping segmentation. On top of a shared encoder-decoder backbone, we utilize multiple branches for semantic segmentation, object detection, and instance center prediction. Finally, our panoptic head combines all outputs into a panoptic segmentation and can even handle conflicting predictions between branches as well as certain false predictions. Our network achieves 32.6% PQ on MS-COCO at 23.5 FPS, opening up panoptic segmentation to a broader field of applications.
This paper addresses the challenging task of video captioning which aims to generate descriptions for video data. Recently, the attention-based encoder-decoder structures have been widely used in video captioning. In existing literature, the attention weights are often built from the information of an individual modality, while, the association relationships between multiple modalities are neglected. Motivated by this observation, we propose a video captioning model with High-Order Cross-Modal Attention (HOCA) where the attention weights are calculated based on the high-order correlation tensor to capture the frame-level cross-modal interaction of different modalities sufficiently. Furthermore, we novelly introduce Low-Rank HOCA which adopts tensor decomposition to reduce the extremely large space requirement of HOCA, leading to a practical and efficient implementation in real-world applications. Experimental results on two benchmark datasets, MSVD and MSR-VTT, show that Low-rank HOCA establishes a new state-of-the-art.
Transformer has shown promising results in many sequence to sequence transformation tasks recently. It utilizes a number of feed-forward self-attention layers to replace the recurrent neural networks (RNN) in attention-based encoder decoder (AED) architecture. Self-attention layer learns temporal dependence by incorporating sinusoidal positional embedding of tokens in a sequence for parallel computing. Quicker iteration speed in training than sequential operation of RNN can be obtained. Deeper layers of the transformer also make it perform better than RNN-based AED. However, this parallelization ability is lost when applying scheduled sampling training. Self-attention with sinusoidal positional embedding may cause performance degradations for longer sequences that have similar acoustic or semantic information at different positions as well. To address these problems, we propose to use parallel scheduled sampling (PSS) and relative positional embedding (RPE) to help the transformer generalize to unseen data. Our proposed methods achieve a 7% relative improvement for short utterances and a 70% relative gain for long utterances on a 10,000-hour Mandarin ASR task.
Taking greedy decoding algorithm as it should be, this work focuses on further strengthening the model itself for Chinese word segmentation (CWS), which results in an even more fast and more accurate CWS model. Our model consists of an attention only stacked encoder and a light enough decoder for the greedy segmentation plus two highway connections for smoother training, in which the encoder is composed of a newly proposed Transformer variant, Gaussian-masked Directional (GD) Transformer, and a biaffine attention scorer. With the effective encoder design, our model only needs to take unigram features for scoring. Our model is evaluated on SIGHAN Bakeoff benchmark datasets. The experimental results show that with the highest segmentation speed, the proposed model achieves new state-of-the-art or comparable performance against strong baselines in terms of strict closed test setting.
The emergence of various intelligent mobile applications demands the deployment of powerful deep learning models at resource-constrained mobile devices. The device-edge co-inference framework provides a promising solution by splitting a neural network at a mobile device and an edge computing server. In order to balance the on-device computation and the communication overhead, the splitting point needs to be carefully picked, while the intermediate feature needs to be compressed before transmission. Existing studies decoupled the design of model splitting, feature compression, and communication, which may lead to excessive resource consumption of the mobile device. In this paper, we introduce an end-to-end architecture, named BottleNet++, that consists of an encoder, a non-trainable channel layer, and a decoder for more efficient feature compression and transmission. The encoder and decoder essentially implement joint source-channel coding via convolutional neural networks (CNNs), while explicitly considering the effect of channel noise. By exploiting the strong sparsity and the fault-tolerant property of the intermediate feature in a deep neural network (DNN), BottleNet++ achieves a much higher compression ratio than existing methods. Furthermore, by providing the channel condition to the encoder as an input, our method enjoys a strong generalization ability in different channel conditions. Compared with merely transmitting intermediate data without feature compression, BottleNet++ achieves up to 64x bandwidth reduction over the additive white Gaussian noise channel and up to 256x bit compression ratio in the binary erasure channel, with less than 2% reduction in accuracy. With a higher compression ratio, BottleNet++ enables splitting a DNN at earlier layers, which leads to up to 3x reduction in on-device computation compared with other compression methods.
Auto Composing is an active and appealing research area in the past few years, and lots of efforts have been put into inventing more robust models to solve this problem. With the fast evolution of deep learning techniques, some deep neural network-based language models are becoming dominant. Notably, the transformer structure has been proven to be very efficient and promising in modeling texts. However, the transformer-based language models usually contain huge number of parameters and the size of the model is usually too large to put in production for some storage limited applications. In this paper, we propose a parameter sharing decoder pair (PSDP), which reduces the number of parameters dramatically and at the same time maintains the capability of generating understandable and reasonable compositions. Works created by the proposed model are presented to demonstrate the effectiveness of the model.
Most RNN-based image captioning models receive supervision on the output words to mimic human captions. Therefore, the hidden states can only receive noisy gradient signals via layers of back-propagation through time, leading to less accurate generated captions. Consequently, we propose a novel framework, Hidden State Guidance (HSG), that matches the hidden states in the caption decoder to those in a teacher decoder trained on an easier task of autoencoding the captions conditioned on the image. During training with the REINFORCE algorithm, the conventional rewards are sentence-based evaluation metrics equally distributed to each generated word, no matter their relevance. HSG provides a word-level reward that helps the model learn better hidden representations. Experimental results demonstrate that HSG clearly outperforms various state-of-the-art caption decoders using either raw images or detected objects as inputs.
This paper describes the ON-TRAC Consortium translation systems developed for the end-to-end model task of IWSLT Evaluation 2019 for the English-to-Portuguese language pair. ON-TRAC Consortium is composed of researchers from three French academic laboratories: LIA (Avignon Universit\'e), LIG (Universit\'e Grenoble Alpes), and LIUM (Le Mans Universit\'e). A single end-to-end model built as a neural encoder-decoder architecture with attention mechanism was used for two primary submissions corresponding to the two EN-PT evaluations sets: (1) TED (MuST-C) and (2) How2. In this paper, we notably investigate impact of pooling heterogeneous corpora for training, impact of target tokenization (characters or BPEs), impact of speech input segmentation and we also compare our best end-to-end model (BLEU of 26.91 on MuST-C and 43.82 on How2 validation sets) to a pipeline (ASR+MT) approach.
A fundamental challenge in artificial intelligence is to build an agent that generalizes and adapts to unseen environments. A common strategy is to build a decoder that takes the context of the unseen new environment as input and generates a policy accordingly. The current paper studies how to build a decoder for the fundamental continuous control task, linear quadratic regulator (LQR), which can model a wide range of real-world physical environments. We present a simple algorithm for this problem, which uses upper confidence bound (UCB) to refine the estimate of the decoder and balance the exploration-exploitation trade-off. Theoretically, our algorithm enjoys a $\widetilde{O}\left(\sqrt{T}\right)$ regret bound in the online setting where $T$ is the number of environments the agent played. This also implies after playing $\widetilde{O}\left(1/\epsilon^2\right)$ environments, the agent is able to transfer the learned knowledge to obtain an $\epsilon$-suboptimal policy for an unseen environment. To our knowledge, this is first provably efficient algorithm to build a decoder in the continuous control setting. While our main focus is theoretical, we also present experiments that demonstrate the effectiveness of our algorithm.
We present BART, a denoising autoencoder for pretraining sequence-to-sequence models. BART is trained by (1) corrupting text with an arbitrary noising function, and (2) learning a model to reconstruct the original text. It uses a standard Tranformer-based neural machine translation architecture which, despite its simplicity, can be seen as generalizing BERT (due to the bidirectional encoder), GPT (with the left-to-right decoder), and many other more recent pretraining schemes. We evaluate a number of noising approaches, finding the best performance by both randomly shuffling the order of the original sentences and using a novel in-filling scheme, where spans of text are replaced with a single mask token. BART is particularly effective when fine tuned for text generation but also works well for comprehension tasks. It matches the performance of RoBERTa with comparable training resources on GLUE and SQuAD, achieves new state-of-the-art results on a range of abstractive dialogue, question answering, and summarization tasks, with gains of up to 6 ROUGE. BART also provides a 1.1 BLEU increase over a back-translation system for machine translation, with only target language pretraining. We also report ablation experiments that replicate other pretraining schemes within the BART framework, to better measure which factors most influence end-task performance.
With the increasing usage of radiograph images as a most common medical imaging system for diagnosis, treatment planning, and clinical studies, it is increasingly becoming a vital factor to use machine learning-based systems to provide reliable information for surgical pre-planning. Segmentation of pelvic bone in radiograph images is a critical preprocessing step for some applications such as automatic pose estimation and disease detection. However, the encoder-decoder style network known as U-Net has demonstrated limited results due to the challenging complexity of the pelvic shapes, especially in severe patients. In this paper, we propose a novel multi-task segmentation method based on Mask R-CNN architecture. For training, the network weights were initialized by large non-medical dataset and fine-tuned with radiograph images. Furthermore, in the training process, augmented data was generated to improve network performance. Our experiments show that Mask R-CNN utilizing multi-task learning, transfer learning, and data augmentation techniques achieve 0.96 DICE coefficient, which significantly outperforms the U-Net. Notably, for a fair comparison, the same transfer learning and data augmentation techniques have been used for U-net training.
We tackle the task of question generation over knowledge bases. Conventional methods for this task neglect two crucial research issues: 1) the given predicate needs to be expressed; 2) the answer to the generated question needs to be definitive. In this paper, we strive toward the above two issues via incorporating diversified contexts and answer-aware loss. Specifically, we propose a neural encoder-decoder model with multi-level copy mechanisms to generate such questions. Furthermore, the answer aware loss is introduced to make generated questions corresponding to more definitive answers. Experiments demonstrate that our model achieves state-of-the-art performance. Meanwhile, such generated question can express the given predicate and correspond to a definitive answer.