PyTorchKR Papers

최신 논문

총 9,642편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2108.13910

A manifold learning perspective on representation learning: Learning decoder and representations without an encoder

Autoencoders are commonly used in representation learning. They consist of an encoder and a decoder, which provide a straightforward way to map n-dimensional data in input space to a lower m-dimensional representation space and back. The decoder itself defines an m-dimensional manifold in input space. Inspired by manifold learning, we show that the decoder can be trained on its own by learning the representations of the training samples along with the decoder weights using gradient descent. A sum-of-squares loss then corresponds to optimizing the manifold to have the smallest Euclidean distance to the training samples, and similarly for other loss functions. We derive expressions for the number of samples needed to specify the encoder and decoder and show that the decoder generally requires much less training samples to be well-specified compared to the encoder. We discuss training of autoencoders in this perspective and relate to previous work in the field that use noisy training examples and other types of regularization. On the natural image data sets MNIST and CIFAR10, we demonstrate that the decoder is much better suited to learn a low-dimensional representation, especially when trained on small data sets. Using simulated gene regulatory data, we further show that the decoder alone leads to better generalization and meaningful representations. Our approach of training the decoder alone facilitates representation learning even on small data sets and can lead to improved training of autoencoders. We hope that the simple analyses presented will also contribute to an improved conceptual understanding of representation learning.

arXiv:2108.13904

Simultaneous Nuclear Instance and Layer Segmentation in Oral Epithelial Dysplasia

Oral epithelial dysplasia (OED) is a pre-malignant histopathological diagnosis given to lesions of the oral cavity. Predicting OED grade or whether a case will transition to malignancy is critical for early detection and appropriate treatment. OED typically begins in the lower third of the epithelium before progressing upwards with grade severity, thus we have suggested that segmenting intra-epithelial layers, in addition to individual nuclei, may enable researchers to evaluate important layer-specific morphological features for grade/malignancy prediction. We present HoVer-Net+, a deep learning framework to simultaneously segment (and classify) nuclei and (intra-)epithelial layers in H&E stained slides from OED cases. The proposed architecture consists of an encoder branch and four decoder branches for simultaneous instance segmentation of nuclei and semantic segmentation of the epithelial layers. We show that the proposed model achieves the state-of-the-art (SOTA) performance in both tasks, with no additional costs when compared to previous SOTA methods for each task. To the best of our knowledge, ours is the first method for simultaneous nuclear instance segmentation and semantic tissue segmentation, with potential for use in computational pathology for other similar simultaneous tasks and for future studies into malignancy prediction.

arXiv:2108.12920공식 코드 · Papers with Code

KO codes: Inventing Nonlinear Encoding and Decoding for Reliable Wireless Communication via Deep-learning

Landmark codes underpin reliable physical layer communication, e.g., Reed-Muller, BCH, Convolution, Turbo, LDPC and Polar codes: each is a linear code and represents a mathematical breakthrough. The impact on humanity is huge: each of these codes has been used in global wireless communication standards (satellite, WiFi, cellular). Reliability of communication over the classical additive white Gaussian noise (AWGN) channel enables benchmarking and ranking of the different codes. In this paper, we construct KO codes, a computationaly efficient family of deep-learning driven (encoder, decoder) pairs that outperform the state-of-the-art reliability performance on the standardized AWGN channel. KO codes beat state-of-the-art Reed-Muller and Polar codes, under the low-complexity successive cancellation decoding, in the challenging short-to-medium block length regime on the AWGN channel. We show that the gains of KO codes are primarily due to the nonlinear mapping of information bits directly to transmit real symbols (bypassing modulation) and yet possess an efficient, high performance decoder. The key technical innovation that renders this possible is design of a novel family of neural architectures inspired by the computation tree of the {\bf K}ronecker {\bf O}peration (KO) central to Reed-Muller and Polar codes. These architectures pave way for the discovery of a much richer class of hitherto unexplored nonlinear algebraic structures. The code is available at \href{https://github.com/deepcomm/KOcodes}{https://github.com/deepcomm/KOcodes}

arXiv:2108.12960공식 코드 · Papers with Code

LOT: A Story-Centric Benchmark for Evaluating Chinese Long Text Understanding and Generation

Standard multi-task benchmarks are essential for developing pretraining models that can generalize to various downstream tasks. Existing benchmarks for natural language processing (NLP) usually focus only on understanding or generating short texts. However, long text modeling requires many distinct abilities in contrast to short texts, such as the modeling of long-range discourse and commonsense relations, and the coherence and controllability of generation. The lack of standardized benchmarks makes it difficult to assess these abilities of a model and fairly compare different models, especially Chinese models. Therefore, we propose a story-centric benchmark named LOT for evaluating Chinese long text modeling, which aggregates two understanding tasks and two generation tasks. We construct new datasets for these tasks based on human-written Chinese stories with hundreds of words. Furthermore, we release an encoder-decoder-based Chinese long text pretraining model named LongLM with up to 1 billion parameters. We pretrain LongLM on 120G Chinese novels with two generative tasks including text infilling and conditional continuation. Extensive experiments show that LongLM outperforms similar-sized pretraining models substantially on both the understanding and generation tasks in LOT.

arXiv:2108.13009

Communication-Computation Efficient Device-Edge Co-Inference via AutoML

Device-edge co-inference, which partitions a deep neural network between a resource-constrained mobile device and an edge server, recently emerges as a promising paradigm to support intelligent mobile applications. To accelerate the inference process, on-device model sparsification and intermediate feature compression are regarded as two prominent techniques. However, as the on-device model sparsity level and intermediate feature compression ratio have direct impacts on computation workload and communication overhead respectively, and both of them affect the inference accuracy, finding the optimal values of these hyper-parameters brings a major challenge due to the large search space. In this paper, we endeavor to develop an efficient algorithm to determine these hyper-parameters. By selecting a suitable model split point and a pair of encoder/decoder for the intermediate feature vector, this problem is casted as a sequential decision problem, for which, a novel automated machine learning (AutoML) framework is proposed based on deep reinforcement learning (DRL). Experiment results on an image classification task demonstrate the effectiveness of the proposed framework in achieving a better communication-computation trade-off and significant inference speedup against various baseline schemes.

arXiv:2108.13004

X2Teeth: 3D Teeth Reconstruction from a Single Panoramic Radiograph

3D teeth reconstruction from X-ray is important for dental diagnosis and many clinical operations. However, no existing work has explored the reconstruction of teeth for a whole cavity from a single panoramic radiograph. Different from single object reconstruction from photos, this task has the unique challenge of constructing multiple objects at high resolutions. To conquer this task, we develop a novel ConvNet X2Teeth that decomposes the task into teeth localization and single-shape estimation. We also introduce a patch-based training strategy, such that X2Teeth can be end-to-end trained for optimal performance. Extensive experiments show that our method can successfully estimate the 3D structure of the cavity and reflect the details for each tooth. Moreover, X2Teeth achieves a reconstruction IoU of 0.681, which significantly outperforms the encoder-decoder method by $1.71X and the retrieval-based method by $1.52X. Our method can also be promising for other multi-anatomy 3D reconstruction tasks.

arXiv:2108.12599공식 코드 · Papers with Code

Smoothing Dialogue States for Open Conversational Machine Reading

Conversational machine reading (CMR) requires machines to communicate with humans through multi-turn interactions between two salient dialogue states of decision making and question generation processes. In open CMR settings, as the more realistic scenario, the retrieved background knowledge would be noisy, which results in severe challenges in the information transmission. Existing studies commonly train independent or pipeline systems for the two subtasks. However, those methods are trivial by using hard-label decisions to activate question generation, which eventually hinders the model performance. In this work, we propose an effective gating strategy by smoothing the two dialogue states in only one decoder and bridge decision making and question generation to provide a richer dialogue state reference. Experiments on the OR-ShARC dataset show the effectiveness of our method, which achieves new state-of-the-art results.

arXiv:2108.12156

Improving callsign recognition with air-surveillance data in air-traffic communication

Automatic Speech Recognition (ASR) can be used as the assistance of speech communication between pilots and air-traffic controllers. Its application can significantly reduce the complexity of the task and increase the reliability of transmitted information. Evidently, high accuracy predictions are needed to minimize the risk of errors. Especially, high accuracy is required in recognition of key information, such as commands and callsigns, used to navigate pilots. Our results prove that the surveillance data containing callsigns can help to considerably improve the recognition of a callsign in an utterance when the weights of probable callsign n-grams are reduced per utterance. In this paper, we investigate two approaches: (1) G-boosting, when callsigns weights are adjusted at language model level (G) and followed by the dynamic decoder with an on-the-fly composition, and (2) lattice rescoring when callsign information is introduced on top of lattices generated using a conventional decoder. Boosting callsign n-grams with the combination of two methods allowed us to gain 28.4% of absolute improvement in callsign recognition accuracy and up to 74.2% of relative improvement in WER of callsign recognition.

arXiv:2108.11992공식 코드 · Papers with Code

Enhanced Seq2Seq Autoencoder via Contrastive Learning for Abstractive Text Summarization

In this paper, we present a denoising sequence-to-sequence (seq2seq) autoencoder via contrastive learning for abstractive text summarization. Our model adopts a standard Transformer-based architecture with a multi-layer bi-directional encoder and an auto-regressive decoder. To enhance its denoising ability, we incorporate self-supervised contrastive learning along with various sentence-level document augmentation. These two components, seq2seq autoencoder and contrastive learning, are jointly trained through fine-tuning, which improves the performance of text summarization with regard to ROUGE scores and human evaluation. We conduct experiments on two datasets and demonstrate that our model outperforms many existing benchmarks and even achieves comparable performance to the state-of-the-art abstractive systems trained with more complex architecture and extensive computation resources.

arXiv:2108.12399

A Novel Hierarchical Light Field Coding Scheme Based on Hybrid Stacked Multiplicative Layers and Fourier Disparity Layers for Glasses-Free 3D Displays

This paper presents a novel hierarchical coding scheme for light fields based on transmittance patterns of low-rank multiplicative layers and Fourier disparity layers. The proposed scheme identifies multiplicative layers of light field view subsets optimized using a convolutional neural network for different scanning orders. Our approach exploits the hidden low-rank structure in the multiplicative layers obtained from the subsets of different scanning patterns. The spatial redundancies in the multiplicative layers can be efficiently removed by performing low-rank approximation at different ranks on the Krylov subspace. The intra-view and inter-view redundancies between approximated layers are further removed by HEVC encoding. Next, a Fourier disparity layer representation is constructed from the first subset of the approximated light field based on the chosen hierarchical order. Subsequent view subsets are synthesized by modeling the Fourier disparity layers that iteratively refine the representation with improved accuracy. The critical advantage of the proposed hybrid layered representation and coding scheme is that it utilizes not just spatial and temporal redundancies in light fields but efficiently exploits intrinsic similarities among neighboring sub-aperture images in both horizontal and vertical directions as specified by different predication orders. In addition, the scheme is flexible to realize a range of multiple bitrates at the decoder within a single integrated system. The compression performance of the proposed scheme is analyzed on real light fields. We achieved substantial bitrate savings and maintained good light field reconstruction quality.

arXiv:2108.11846공식 코드 · Papers with Code

Alleviating Exposure Bias via Contrastive Learning for Abstractive Text Summarization

Encoder-decoder models have achieved remarkable success in abstractive text summarization, which aims to compress one or more documents into a shorter version without the loss of the essential content. Unfortunately, these models mostly suffer a discrepancy between training and inference, i.e., the exposure bias problem. During the training stage, with teacher forcing these models are optimized to maximize the likelihood of the gold summary given the gold summary tokens as input to the decoder, while at inference the given tokens are replaced by the generated tokens. Consequently, low-quality summaries are very likely to be generated. To remedy this problem, we propose to leverage contrastive learning to decrease the likelihood of these low-quality summaries, and meanwhile increase the likelihood of the gold summary. Since our solution expands the states that the model perceives during training, we expect that the exposure bias problem can be alleviated. We experimentally demonstrate that our method effectively improves the performance of the state-of-the-art model on different datasets.

arXiv:2108.11763

Attention-based Neural Load Forecasting: A Dynamic Feature Selection Approach

Encoder-decoder-based recurrent neural network (RNN) has made significant progress in sequence-to-sequence learning tasks such as machine translation and conversational models. Recent works have shown the advantage of this type of network in dealing with various time series forecasting tasks. The present paper focuses on the problem of multi-horizon short-term load forecasting, which plays a key role in the power system's planning and operation. Leveraging the encoder-decoder RNN, we develop an attention model to select the relevant features and similar temporal information adaptively. First, input features are assigned with different weights by a feature selection attention layer, while the updated historical features are encoded by a bi-directional long short-term memory (BiLSTM) layer. Then, a decoder with hierarchical temporal attention enables a similar day selection, which re-evaluates the importance of historical information at each time step. Numerical results tested on the dataset of the global energy forecasting competition 2014 show that our proposed model significantly outperforms some existing forecasting schemes.

arXiv:2108.11730

Deep learning based dictionary learning and tomographic image reconstruction

This work presents an approach for image reconstruction in clinical low-dose tomography that combines principles from sparse signal processing with ideas from deep learning. First, we describe sparse signal representation in terms of dictionaries from a statistical perspective and interpret dictionary learning as a process of aligning distribution that arises from a generative model with empirical distribution of true signals. As a result we can see that sparse coding with learned dictionaries resembles a specific variational autoencoder, where the decoder is a linear function and the encoder is a sparse coding algorithm. Next, we show that dictionary learning can also benefit from computational advancements introduced in the context of deep learning, such as parallelism and as stochastic optimization. Finally, we show that regularization by dictionaries achieves competitive performance in computed tomography (CT) reconstruction comparing to state-of-the-art model based and data driven approaches.

arXiv:2108.11609

Unsupervised Dense Deformation Embedding Network for Template-Free Shape Correspondence

Shape correspondence from 3D deformation learning has attracted appealing academy interests recently. Nevertheless, current deep learning based methods require the supervision of dense annotations to learn per-point translations, which severely overparameterize the deformation process. Moreover, they fail to capture local geometric details of original shape via global feature embedding. To address these challenges, we develop a new Unsupervised Dense Deformation Embedding Network (i.e., UD^2E-Net), which learns to predict deformations between non-rigid shapes from dense local features. Since it is non-trivial to match deformation-variant local features for deformation prediction, we develop an Extrinsic-Intrinsic Autoencoder to frst encode extrinsic geometric features from source into intrinsic coordinates in a shared canonical shape, with which the decoder then synthesizes corresponding target features. Moreover, a bounded maximum mean discrepancy loss is developed to mitigate the distribution divergence between the synthesized and original features. To learn natural deformation without dense supervision, we introduce a coarse parameterized deformation graph, for which a novel trace and propagation algorithm is proposed to improve both the quality and effciency of the deformation. Our UD^2E-Net outperforms state-of-the-art unsupervised methods by 24% on Faust Inter challenge and even supervised methods by 13% on Faust Intra challenge.

arXiv:2012.08364공식 코드 · Papers with Code

GAP-net for Snapshot Compressive Imaging

Snapshot compressive imaging (SCI) systems aim to capture high-dimensional ($\ge3$D) images in a single shot using 2D detectors. SCI devices include two main parts: a hardware encoder and a software decoder. The hardware encoder typically consists of an (optical) imaging system designed to capture {compressed measurements}. The software decoder on the other hand refers to a reconstruction algorithm that retrieves the desired high-dimensional signal from those measurements. In this paper, using deep unfolding ideas, we propose an SCI recovery algorithm, namely GAP-net, which unfolds the generalized alternating projection (GAP) algorithm. At each stage, GAP-net passes its current estimate of the desired signal through a trained convolutional neural network (CNN). The CNN operates as a denoiser that projects the estimate back to the desired signal space. For the GAP-net that employs trained auto-encoder-based denoisers, we prove a probabilistic global convergence result. Finally, we investigate the performance of GAP-net in solving video SCI and spectral SCI problems. In both cases, GAP-net demonstrates competitive performance on both synthetic and real data. In addition to having high accuracy and high speed, we show that GAP-net is flexible with respect to signal modulation implying that a trained GAP-net decoder can be applied in different systems. Our code is at https://github.com/mengziyi64/ADMM-net.

arXiv:2108.11244

Multiscale Spatio-Temporal Graph Neural Networks for 3D Skeleton-Based Motion Prediction

We propose a multiscale spatio-temporal graph neural network (MST-GNN) to predict the future 3D skeleton-based human poses in an action-category-agnostic manner. The core of MST-GNN is a multiscale spatio-temporal graph that explicitly models the relations in motions at various spatial and temporal scales. Different from many previous hierarchical structures, our multiscale spatio-temporal graph is built in a data-adaptive fashion, which captures nonphysical, yet motion-based relations. The key module of MST-GNN is a multiscale spatio-temporal graph computational unit (MST-GCU) based on the trainable graph structure. MST-GCU embeds underlying features at individual scales and then fuses features across scales to obtain a comprehensive representation. The overall architecture of MST-GNN follows an encoder-decoder framework, where the encoder consists of a sequence of MST-GCUs to learn the spatial and temporal features of motions, and the decoder uses a graph-based attention gate recurrent unit (GA-GRU) to generate future poses. Extensive experiments are conducted to show that the proposed MST-GNN outperforms state-of-the-art methods in both short and long-term motion prediction on the datasets of Human 3.6M, CMU Mocap and 3DPW, where MST-GNN outperforms previous works by 5.33% and 3.67% of mean angle errors in average for short-term and long-term prediction on Human 3.6M, and by 11.84% and 4.71% of mean angle errors for short-term and long-term prediction on CMU Mocap, and by 1.13% of mean angle errors on 3DPW in average, respectively. We further investigate the learned multiscale graphs for interpretability.

arXiv:2108.10764

Regularizing Transformers With Deep Probabilistic Layers

Language models (LM) have grown with non-stop in the last decade, from sequence-to-sequence architectures to the state-of-the-art and utter attention-based Transformers. In this work, we demonstrate how the inclusion of deep generative models within BERT can bring more versatile models, able to impute missing/noisy words with richer text or even improve BLEU score. More precisely, we use a Gaussian Mixture Variational Autoencoder (GMVAE) as a regularizer layer and prove its effectiveness not only in Transformers but also in the most relevant encoder-decoder based LM, seq2seq with and without attention.

arXiv:2108.10506공식 코드 · Papers with Code

Real-Time Monocular Human Depth Estimation and Segmentation on Embedded Systems

Estimating a scene's depth to achieve collision avoidance against moving pedestrians is a crucial and fundamental problem in the robotic field. This paper proposes a novel, low complexity network architecture for fast and accurate human depth estimation and segmentation in indoor environments, aiming to applications for resource-constrained platforms (including battery-powered aerial, micro-aerial, and ground vehicles) with a monocular camera being the primary perception module. Following the encoder-decoder structure, the proposed framework consists of two branches, one for depth prediction and another for semantic segmentation. Moreover, network structure optimization is employed to improve its forward inference speed. Exhaustive experiments on three self-generated datasets prove our pipeline's capability to execute in real-time, achieving higher frame rates than contemporary state-of-the-art frameworks (114.6 frames per second on an NVIDIA Jetson Nano GPU with TensorRT) while maintaining comparable accuracy.

arXiv:2108.10417

Recurrent multiple shared layers in Depth for Neural Machine Translation

Learning deeper models is usually a simple and effective approach to improve model performance, but deeper models have larger model parameters and are more difficult to train. To get a deeper model, simply stacking more layers of the model seems to work well, but previous works have claimed that it cannot benefit the model. We propose to train a deeper model with recurrent mechanism, which loops the encoder and decoder blocks of Transformer in the depth direction. To address the increasing of model parameters, we choose to share parameters in different recursive moments. We conduct our experiments on WMT16 English-to-German and WMT14 English-to-France translation tasks, our model outperforms the shallow Transformer-Base/Big baseline by 0.35, 1.45 BLEU points, which is 27.23% of Transformer-Big model parameters. Compared to the deep Transformer(20-layer encoder, 6-layer decoder), our model has similar model performance and infer speed, but our model parameters are 54.72% of the former.

arXiv:2108.10531공식 코드 · Papers with Code

Unsupervised Depth Completion with Calibrated Backprojection Layers

We propose a deep neural network architecture to infer dense depth from an image and a sparse point cloud. It is trained using a video stream and corresponding synchronized sparse point cloud, as obtained from a LIDAR or other range sensor, along with the intrinsic calibration parameters of the camera. At inference time, the calibration of the camera, which can be different than the one used for training, is fed as an input to the network along with the sparse point cloud and a single image. A Calibrated Backprojection Layer backprojects each pixel in the image to three-dimensional space using the calibration matrix and a depth feature descriptor. The resulting 3D positional encoding is concatenated with the image descriptor and the previous layer output to yield the input to the next layer of the encoder. A decoder, exploiting skip-connections, produces a dense depth map. The resulting Calibrated Backprojection Network, or KBNet, is trained without supervision by minimizing the photometric reprojection error. KBNet imputes missing depth value based on the training set, rather than on generic regularization. We test KBNet on public depth completion benchmarks, where it outperforms the state of the art by 30.5% indoor and 8.8% outdoor when the same camera is used for training and testing. When the test camera is different, the improvement reaches 62%. Code available at: https://github.com/alexklwong/calibrated-backprojection-network.