Atomic clauses are fundamental text units for understanding complex sentences. Identifying the atomic sentences within complex sentences is important for applications such as summarization, argument mining, discourse analysis, discourse parsing, and question answering. Previous work mainly relies on rule-based methods dependent on parsing. We propose a new task to decompose each complex sentence into simple sentences derived from the tensed clauses in the source, and a novel problem formulation as a graph edit task. Our neural model learns to Accept, Break, Copy or Drop elements of a graph that combines word adjacency and grammatical dependencies. The full processing pipeline includes modules for graph construction, graph editing, and sentence generation from the output graph. We introduce DeSSE, a new dataset designed to train and evaluate complex sentence decomposition, and MinWiki, a subset of MinWikiSplit. ABCD achieves comparable performance as two parsing baselines on MinWiki. On DeSSE, which has a more even balance of complex sentence types, our model achieves higher accuracy on the number of atomic sentences than an encoder-decoder baseline. Results include a detailed error analysis.
We present a novel deep neural network (DNN) architecture for compressing an image when a correlated image is available as side information only at the decoder. This problem is known as distributed source coding (DSC) in information theory. In particular, we consider a pair of stereo images, which generally have high correlation with each other due to overlapping fields of view, and assume that one image of the pair is to be compressed and transmitted, while the other image is available only at the decoder. In the proposed architecture, the encoder maps the input image to a latent space, quantizes the latent representation, and compresses it using entropy coding. The decoder is trained to extract the common information between the input image and the correlated image, using only the latter. The received latent representation and the locally generated common information are passed through a decoder network to obtain an enhanced reconstruction of the input image. The common information provides a succinct representation of the relevant information at the receiver. We train and demonstrate the effectiveness of the proposed approach on the KITTI and Cityscape datasets of stereo image pairs. Our results show that the proposed architecture is capable of exploiting the decoder-only side information, and outperforms previous work on stereo image compression with decoder side information.
Graph representation learning is a fundamental problem for modeling relational data and benefits a number of downstream applications. Traditional Bayesian-based graph models and recent deep learning based GNN either suffer from impracticability or lack interpretability, thus combined models for undirected graphs have been proposed to overcome the weaknesses. As a large portion of real-world graphs are directed graphs (of which undirected graphs are special cases), in this paper, we propose a Deep Latent Space Model (DLSM) for directed graphs to incorporate the traditional latent variable based generative model into deep learning frameworks. Our proposed model consists of a graph convolutional network (GCN) encoder and a stochastic decoder, which are layer-wise connected by a hierarchical variational auto-encoder architecture. By specifically modeling the degree heterogeneity using node random factors, our model possesses better interpretability in both community structure and degree heterogeneity. For fast inference, the stochastic gradient variational Bayes (SGVB) is adopted using a non-iterative recognition model, which is much more scalable than traditional MCMC-based methods. The experiments on real-world datasets show that the proposed model achieves the state-of-the-art performances on both link prediction and community detection tasks while learning interpretable node embeddings. The source code is available at https://github.com/upperr/DLSM.
A wide variety of NLP applications, such as machine translation, summarization, and dialog, involve text generation. One major challenge for these applications is how to evaluate whether such generated texts are actually fluent, accurate, or effective. In this work, we conceptualize the evaluation of generated text as a text generation problem, modeled using pre-trained sequence-to-sequence models. The general idea is that models trained to convert the generated text to/from a reference output or the source text will achieve higher scores when the generated text is better. We operationalize this idea using BART, an encoder-decoder based pre-trained model, and propose a metric BARTScore with a number of variants that can be flexibly applied in an unsupervised fashion to evaluation of text from different perspectives (e.g. informativeness, fluency, or factuality). BARTScore is conceptually simple and empirically effective. It can outperform existing top-scoring metrics in 16 of 22 test settings, covering evaluation of 16 datasets (e.g., machine translation, text summarization) and 7 different perspectives (e.g., informativeness, factuality). Code to calculate BARTScore is available at https://github.com/neulab/BARTScore, and we have released an interactive leaderboard for meta-evaluation at http://explainaboard.nlpedia.ai/leaderboard/task-meval/ on the ExplainaBoard platform, which allows us to interactively understand the strengths, weaknesses, and complementarity of each metric.
Coronary X-ray angiography is a crucial clinical procedure for the diagnosis and treatment of coronary artery disease, which accounts for roughly 16% of global deaths every year. However, the images acquired in these procedures have low resolution and poor contrast, making lesion detection and assessment challenging. Accurate coronary artery segmentation not only helps mitigate these problems, but also allows the extraction of relevant anatomical features for further analysis by quantitative methods. Although automated segmentation of coronary arteries has been proposed before, previous approaches have used non-optimal segmentation criteria, leading to less useful results. Most methods either segment only the major vessel, discarding important information from the remaining ones, or segment the whole coronary tree based mostly on contrast information, producing a noisy output that includes vessels that are not relevant for diagnosis. We adopt a better-suited clinical criterion and segment vessels according to their clinical relevance. Additionally, we simultaneously perform catheter segmentation, which may be useful for diagnosis due to the scale factor provided by the catheter's known diameter, and is a task that has not yet been performed with good results. To derive the optimal approach, we conducted an extensive comparative study of encoder-decoder architectures trained on a combination of focal loss and a variant of generalized dice loss. Based on the EfficientNet and the UNet++ architectures, we propose a line of efficient and high-performance segmentation models using a new decoder architecture, the EfficientUNet++, whose best-performing version achieved average dice scores of 0.8904 and 0.7526 for the artery and catheter classes, respectively, and an average generalized dice score of 0.9234.
In recent years, the size of pre-trained language models (PLMs) has grown by leaps and bounds. However, efficiency issues of these large-scale PLMs limit their utilization in real-world scenarios. We present a suite of cost-effective techniques for the use of PLMs to deal with the efficiency issues of pre-training, fine-tuning, and inference. (1) We introduce knowledge inheritance to accelerate the pre-training process by exploiting existing PLMs instead of training models from scratch. (2) We explore the best practice of prompt tuning with large-scale PLMs. Compared with conventional fine-tuning, prompt tuning significantly reduces the number of task-specific parameters. (3) We implement a new inference toolkit, namely InfMoE, for using large-scale PLMs with limited computational resources. Based on our cost-effective pipeline, we pre-train two models: an encoder-decoder bilingual model with 11 billion parameters (CPM-2) and its corresponding MoE version with 198 billion parameters. In our experiments, we compare CPM-2 with mT5 on downstream tasks. Experimental results show that CPM-2 has excellent general language intelligence. Moreover, we validate the efficiency of InfMoE when conducting inference of large-scale models having tens of billions of parameters on a single GPU. All source code and model parameters are available at https://github.com/TsinghuaAI/CPM.
Most recent approaches for online action detection tend to apply Recurrent Neural Network (RNN) to capture long-range temporal structure. However, RNN suffers from non-parallelism and gradient vanishing, hence it is hard to be optimized. In this paper, we propose a new encoder-decoder framework based on Transformers, named OadTR, to tackle these problems. The encoder attached with a task token aims to capture the relationships and global interactions between historical observations. The decoder extracts auxiliary information by aggregating anticipated future clip representations. Therefore, OadTR can recognize current actions by encoding historical information and predicting future context simultaneously. We extensively evaluate the proposed OadTR on three challenging datasets: HDD, TVSeries, and THUMOS14. The experimental results show that OadTR achieves higher training and inference speeds than current RNN based approaches, and significantly outperforms the state-of-the-art methods in terms of both mAP and mcAP. Code is available at https://github.com/wangxiang1230/OadTR.
In this paper, we propose SPBERT, a transformer-based language model pre-trained on massive SPARQL query logs. By incorporating masked language modeling objectives and the word structural objective, SPBERT can learn general-purpose representations in both natural language and SPARQL query language. We investigate how SPBERT and encoder-decoder architecture can be adapted for Knowledge-based QA corpora. We conduct exhaustive experiments on two additional tasks, including SPARQL Query Construction and Answer Verbalization Generation. The experimental results show that SPBERT can obtain promising results, achieving state-of-the-art BLEU scores on several of these tasks.
Quotations are crucial for successful explanations and persuasions in interpersonal communications. However, finding what to quote in a conversation is challenging for both humans and machines. This work studies automatic quotation generation in an online conversation and explores how language consistency affects whether a quotation fits the given context. Here, we capture the contextual consistency of a quotation in terms of latent topics, interactions with the dialogue history, and coherence to the query turn's existing content. Further, an encoder-decoder neural framework is employed to continue the context with a quotation via language generation. Experiment results on two large-scale datasets in English and Chinese demonstrate that our quotation generation model outperforms the state-of-the-art models. Further analysis shows that topic, interaction, and query consistency are all helpful to learn how to quote in online conversations.
In this work we address the problem of landmark-based segmentation for anatomical structures. We propose HybridGNet, an encoder-decoder neural architecture which combines standard convolutions for image feature encoding, with graph convolutional neural networks to decode plausible representations of anatomical structures. We benchmark the proposed architecture considering other standard landmark and pixel-based models for anatomical segmentation in chest x-ray images, and found that HybridGNet is more robust to image occlusions. We also show that it can be used to construct landmark-based segmentations from pixel level annotations. Our experimental results suggest that HybridGNet produces accurate and anatomically plausible landmark-based segmentations, by naturally incorporating shape constraints within the decoding process via spectral convolutions.
Voice Conversion (VC) is a technique that aims to transform the non-linguistic information of a source utterance to change the perceived identity of the speaker. While there is a rich literature on VC, most proposed methods are trained and evaluated on clean speech recordings. However, many acoustic environments are noisy and reverberant, severely restricting the applicability of popular VC methods to such scenarios. To address this limitation, we propose Voicy, a new VC framework particularly tailored for noisy speech. Our method, which is inspired by the de-noising auto-encoders framework, is comprised of four encoders (speaker, content, phonetic and acoustic-ASR) and one decoder. Importantly, Voicy is capable of performing non-parallel zero-shot VC, an important requirement for any VC system that needs to work on speakers not seen during training. We have validated our approach using a noisy reverberant version of the LibriSpeech dataset. Experimental results show that Voicy outperforms other tested VC techniques in terms of naturalness and target speaker similarity in noisy reverberant environments.
In design, fabrication, and control problems, we are often faced with the task of synthesis, in which we must generate an object or configuration that satisfies a set of constraints while maximizing one or more objective functions. The synthesis problem is typically characterized by a physical process in which many different realizations may achieve the goal. This many-to-one map presents challenges to the supervised learning of feed-forward synthesis, as the set of viable designs may have a complex structure. In addition, the non-differentiable nature of many physical simulations prevents efficient direct optimization. We address both of these problems with a two-stage neural network architecture that we may consider to be an autoencoder. We first learn the decoder: a differentiable surrogate that approximates the many-to-one physical realization process. We then learn the encoder, which maps from goal to design, while using the fixed decoder to evaluate the quality of the realization. We evaluate the approach on two case studies: extruder path planning in additive manufacturing and constrained soft robot inverse kinematics. We compare our approach to direct optimization of the design using the learned surrogate, and to supervised learning of the synthesis problem. We find that our approach produces higher quality solutions than supervised learning, while being competitive in quality with direct optimization, at a greatly reduced computational cost.
The cyclically equivariant neural decoder was recently proposed in [Chen-Ye, International Conference on Machine Learning, 2021] to decode cyclic codes. In the same paper, a list decoding procedure was also introduced for two widely used classes of cyclic codes -- BCH codes and punctured Reed-Muller (RM) codes. While the list decoding procedure significantly improves the Frame Error Rate (FER) of the cyclically equivariant neural decoder, the Bit Error Rate (BER) of the list decoding procedure is even worse than the unique decoding algorithm when the list size is small. In this paper, we propose an improved version of the list decoding algorithm for BCH codes and punctured RM codes. Our new proposal significantly reduces the BER while maintaining the same (in some cases even smaller) FER. More specifically, our new decoder provides up to $2$dB gain over the previous list decoder when measured by BER, and the running time of our new decoder is $15\%$ smaller. Code available at https://github.com/improvedlistdecoder/code
While multitask and transfer learning has shown to improve the performance of neural networks in limited data settings, they require pretraining of the model on large datasets beforehand. In this paper, we focus on improving the performance of weakly supervised sound event detection in low data and noisy settings simultaneously without requiring any pretraining task. To that extent, we propose a shared encoder architecture with sound event detection as a primary task and an additional secondary decoder for a self-supervised auxiliary task. We empirically evaluate the proposed framework for weakly supervised sound event detection on a remix dataset of the DCASE 2019 task 1 acoustic scene data with DCASE 2018 Task 2 sounds event data under 0, 10 and 20 dB SNR. To ensure we retain the localisation information of multiple sound events, we propose a two-step attention pooling mechanism that provides a time-frequency localisation of multiple audio events in the clip. The proposed framework with two-step attention outperforms existing benchmark models by 22.3%, 12.8%, 5.9% on 0, 10 and 20 dB SNR respectively. We carry out an ablation study to determine the contribution of the auxiliary task and two-step attention pooling to the SED performance improvement.
Automatic medical image segmentation has made great progress benefit from the development of deep learning. However, most existing methods are based on convolutional neural networks (CNNs), which fail to build long-range dependencies and global context connections due to the limitation of receptive field in convolution operation. Inspired by the success of Transformer in modeling the long-range contextual information, some researchers have expended considerable efforts in designing the robust variants of Transformer-based U-Net. Moreover, the patch division used in vision transformers usually ignores the pixel-level intrinsic structural features inside each patch. To alleviate these problems, we propose a novel deep medical image segmentation framework called Dual Swin Transformer U-Net (DS-TransUNet), which might be the first attempt to concurrently incorporate the advantages of hierarchical Swin Transformer into both encoder and decoder of the standard U-shaped architecture to enhance the semantic segmentation quality of varying medical images. Unlike many prior Transformer-based solutions, the proposed DS-TransUNet first adopts dual-scale encoder subnetworks based on Swin Transformer to extract the coarse and fine-grained feature representations of different semantic scales. As the core component for our DS-TransUNet, a well-designed Transformer Interactive Fusion (TIF) module is proposed to effectively establish global dependencies between features of different scales through the self-attention mechanism. Furthermore, we also introduce the Swin Transformer block into decoder to further explore the long-range contextual information during the up-sampling process. Extensive experiments across four typical tasks for medical image segmentation demonstrate the effectiveness of DS-TransUNet, and show that our approach significantly outperforms the state-of-the-art methods.
Attention-based encoder-decoder framework is widely used in the scene text recognition task. However, for the current state-of-the-art(SOTA) methods, there is room for improvement in terms of the efficient usage of local visual and global context information of the input text image, as well as the robust correlation between the scene processing module(encoder) and the text processing module(decoder). In this paper, we propose a Representation and Correlation Enhanced Encoder-Decoder Framework(RCEED) to address these deficiencies and break performance bottleneck. In the encoder module, local visual feature, global context feature, and position information are aligned and fused to generate a small-size comprehensive feature map. In the decoder module, two methods are utilized to enhance the correlation between scene and text feature space. 1) The decoder initialization is guided by the holistic feature and global glimpse vector exported from the encoder. 2) The feature enriched glimpse vector produced by the Multi-Head General Attention is used to assist the RNN iteration and the character prediction at each time step. Meanwhile, we also design a Layernorm-Dropout LSTM cell to improve model's generalization towards changeable texts. Extensive experiments on the benchmarks demonstrate the advantageous performance of RCEED in scene text recognition tasks, especially the irregular ones.
Deep learning has become the dominant approach in coping with various tasks in Natural LanguageProcessing (NLP). Although text inputs are typically represented as a sequence of tokens, there isa rich variety of NLP problems that can be best expressed with a graph structure. As a result, thereis a surge of interests in developing new deep learning techniques on graphs for a large numberof NLP tasks. In this survey, we present a comprehensive overview onGraph Neural Networks(GNNs) for Natural Language Processing. We propose a new taxonomy of GNNs for NLP, whichsystematically organizes existing research of GNNs for NLP along three axes: graph construction,graph representation learning, and graph based encoder-decoder models. We further introducea large number of NLP applications that are exploiting the power of GNNs and summarize thecorresponding benchmark datasets, evaluation metrics, and open-source codes. Finally, we discussvarious outstanding challenges for making the full use of GNNs for NLP as well as future researchdirections. To the best of our knowledge, this is the first comprehensive overview of Graph NeuralNetworks for Natural Language Processing.
This paper describes a novel lossless point cloud compression algorithm that uses a neural network for estimating the coding probabilities for the occupancy status of voxels, depending on wide three dimensional contexts around the voxel to be encoded. The point cloud is represented as an octree, with each resolution layer being sequentially encoded and decoded using arithmetic coding, starting from the lowest resolution, until the final resolution is reached. The occupancy probability of each voxel of the splitting pattern at each node of the octree is modeled by a neural network, having at its input the already encoded occupancy status of several octree nodes (belonging to the past and current resolutions), corresponding to a 3D context surrounding the node to be encoded. The algorithm has a fast and a slow version, the fast version selecting differently several voxels of the context, which allows an increased parallelization by sending larger batches of templates to be estimated by the neural network, at both encoder and decoder. The proposed algorithms yield state-of-the-art results on benchmark datasets. The implementation will be made available at https://github.com/marmus12/nnctx
Unsupervised representation learning has recently received lots of interest due to its powerful generalizability through effectively leveraging large-scale unlabeled data. There are two prevalent approaches for this, contrastive learning and generative pre-training, where the former learns representations from instance-wise discrimination tasks and the latter learns them from estimating the likelihood. These seemingly orthogonal approaches have their own strengths and weaknesses. Contrastive learning tends to extract semantic information and discards details irrelevant for classifying objects, making the representations effective for discriminative tasks while degrading robustness to out-of-distribution data. On the other hand, the generative pre-training directly estimates the data distribution, so the representations tend to be robust but not optimal for discriminative tasks. In this paper, we show that we could achieve the best of both worlds by a hybrid training scheme. Specifically, we demonstrated that a transformer-based encoder-decoder architecture trained with both contrastive and generative losses can learn highly discriminative and robust representations without hurting the generative performance. We extensively validate our approach on various tasks.
Maintaining consistent personas is essential for dialogue agents. Although tremendous advancements have been brought, the limited-scale of annotated persona-dense data are still barriers towards training robust and consistent persona-based dialogue models. In this work, we show how the challenges can be addressed by disentangling persona-based dialogue generation into two sub-tasks with a novel BERT-over-BERT (BoB) model. Specifically, the model consists of a BERT-based encoder and two BERT-based decoders, where one decoder is for response generation, and another is for consistency understanding. In particular, to learn the ability of consistency understanding from large-scale non-dialogue inference data, we train the second decoder in an unlikelihood manner. Under different limited data settings, both automatic and human evaluations demonstrate that the proposed model outperforms strong baselines in response quality and persona consistency.