Deep encoder-decoder based CNNs have advanced image inpainting methods for hole filling. While existing methods recover structures and textures step-by-step in the hole regions, they typically use two encoder-decoders for separate recovery. The CNN features of each encoder are learned to capture either missing structures or textures without considering them as a whole. The insufficient utilization of these encoder features limit the performance of recovering both structures and textures. In this paper, we propose a mutual encoder-decoder CNN for joint recovery of both. We use CNN features from the deep and shallow layers of the encoder to represent structures and textures of an input image, respectively. The deep layer features are sent to a structure branch and the shallow layer features are sent to a texture branch. In each branch, we fill holes in multiple scales of the CNN features. The filled CNN features from both branches are concatenated and then equalized. During feature equalization, we reweigh channel attentions first and propose a bilateral propagation activation function to enable spatial equalization. To this end, the filled CNN features of structure and texture mutually benefit each other to represent image content at all feature levels. We use the equalized feature to supplement decoder features for output image generation through skip connections. Experiments on the benchmark datasets show the proposed method is effective to recover structures and textures and performs favorably against state-of-the-art approaches.
Smooth and seamless robot navigation while interacting with humans depends on predicting human movements. Forecasting such human dynamics often involves modeling human trajectories (global motion) or detailed body joint movements (local motion). Prior work typically tackled local and global human movements separately. In this paper, we propose a novel framework to tackle both tasks of human motion (or trajectory) and body skeleton pose forecasting in a unified end-to-end pipeline. To deal with this real-world problem, we consider incorporating both scene and social contexts, as critical clues for this prediction task, into our proposed framework. To this end, we first couple these two tasks by i) encoding their history using a shared Gated Recurrent Unit (GRU) encoder and ii) applying a metric as loss, which measures the source of errors in each task jointly as a single distance. Then, we incorporate the scene context by encoding a spatio-temporal representation of the video data. We also include social clues by generating a joint feature representation from motion and pose of all individuals from the scene using a social pooling layer. Finally, we use a GRU based decoder to forecast both motion and skeleton pose. We demonstrate that our proposed framework achieves a superior performance compared to several baselines on two social datasets.
Existing RGB-D salient object detection (SOD) approaches concentrate on the cross-modal fusion between the RGB stream and the depth stream. They do not deeply explore the effect of the depth map itself. In this work, we design a single stream network to directly use the depth map to guide early fusion and middle fusion between RGB and depth, which saves the feature encoder of the depth stream and achieves a lightweight and real-time model. We tactfully utilize depth information from two perspectives: (1) Overcoming the incompatibility problem caused by the great difference between modalities, we build a single stream encoder to achieve the early fusion, which can take full advantage of ImageNet pre-trained backbone model to extract rich and discriminative features. (2) We design a novel depth-enhanced dual attention module (DEDA) to efficiently provide the fore-/back-ground branches with the spatially filtered features, which enables the decoder to optimally perform the middle fusion. Besides, we put forward a pyramidally attended feature extraction module (PAFE) to accurately localize the objects of different scales. Extensive experiments demonstrate that the proposed model performs favorably against most state-of-the-art methods under different evaluation metrics. Furthermore, this model is 55.5\% lighter than the current lightest model and runs at a real-time speed of 32 FPS when processing a $384 \times 384$ image.
Multi-label classification is the challenging task of predicting the presence and absence of multiple targets, involving representation learning and label correlation modeling. We propose a novel framework for multi-label classification, Multivariate Probit Variational AutoEncoder (MPVAE), that effectively learns latent embedding spaces as well as label correlations. MPVAE learns and aligns two probabilistic embedding spaces for labels and features respectively. The decoder of MPVAE takes in the samples from the embedding spaces and models the joint distribution of output targets under a Multivariate Probit model by learning a shared covariance matrix. We show that MPVAE outperforms the existing state-of-the-art methods on a variety of application domains, using public real-world datasets. MPVAE is further shown to remain robust under noisy settings. Lastly, we demonstrate the interpretability of the learned covariance by a case study on a bird observation dataset.
Generating longer textual sequences when conditioned on the visual information is an interesting problem to explore. The challenge here proliferate over the standard vision conditioned sentence-level generation (e.g., image or video captioning) as it requires to produce a brief and coherent story describing the visual content. In this paper, we mask this Vision-to-Sequence as Graph-to-Sequence learning problem and approach it with the Transformer architecture. To be specific, we introduce Sparse Graph-to-Sequence Transformer (SGST) for encoding the graph and decoding a sequence. The encoder aims to directly encode graph-level semantics, while the decoder is used to generate longer sequences. Experiments conducted with the benchmark image paragraph dataset show that our proposed achieve 13.3% improvement on the CIDEr evaluation measure when comparing to the previous state-of-the-art approach.
Human motion prediction, which aims at predicting future human skeletons given the past ones, is a typical sequence-to-sequence problem. Therefore, extensive efforts have been continued on exploring different RNN-based encoder-decoder architectures. However, by generating target poses conditioned on the previously generated ones, these models are prone to bringing issues such as error accumulation problem. In this paper, we argue that such issue is mainly caused by adopting autoregressive manner. Hence, a novel Non-auToregressive Model (NAT) is proposed with a complete non-autoregressive decoding scheme, as well as a context encoder and a positional encoding module. More specifically, the context encoder embeds the given poses from temporal and spatial perspectives. The frame decoder is responsible for predicting each future pose independently. The positional encoding module injects positional signal into the model to indicate temporal order. Moreover, a multitask training paradigm is presented for both low-level human skeleton prediction and high-level human action recognition, resulting in the convincing improvement for the prediction task. Our approach is evaluated on Human3.6M and CMU-Mocap benchmarks and outperforms state-of-the-art autoregressive methods.
Objective. In this paper, we consider the problem of cross-subject decoding, where neural activity data collected from the prefrontal cortex of a given subject (destination) is used to decode motor intentions from the neural activity of a different subject (source). Approach. We cast the problem of neural activity mapping in a probabilistic framework where we adopt deep generative modelling. Our proposed algorithm uses deep conditional variational autoencoder to infer the representation of the neural activity of the source subject into an adequate feature space of the destination subject where neural decoding takes place. Results. We verify our approach on an experimental data set in which two macaque monkeys perform memory-guided visual saccades to one of eight target locations. The results show a peak cross-subject decoding improvement of $8\%$ over subject-specific decoding. Conclusion. We demonstrate that a neural decoder trained on neural activity signals of one subject can be used to robustly decode the motor intentions of a different subject with high reliability. This is achieved in spite of the non-stationary nature of neural activity signals and the subject-specific variations of the recording conditions. Significance. The findings reported in this paper are an important step towards the development of cross-subject brain-computer that generalize well across a population.
The multivariate time series generated from merchant transaction history can provide critical insights for payment processing companies. The capability of predicting merchants' future is crucial for fraud detection and recommendation systems. Conventionally, this problem is formulated to predict one multivariate time series under the multi-horizon setting. However, real-world applications often require more than one future trend prediction considering the uncertainties, where more than one multivariate time series needs to be predicted. This problem is called multi-future prediction. In this work, we combine the two research directions and propose to study this new problem: multi-future, multi-horizon and multivariate time series prediction. This problem is crucial as it has broad use cases in the financial industry to reduce the risk while improving user experience by providing alternative futures. This problem is also challenging as now we not only need to capture the patterns and insights from the past but also train a model that has a strong inference capability to project multiple possible outcomes. To solve this problem, we propose a new model using convolutional neural networks and a simple yet effective encoder-decoder structure to learn the time series pattern from multiple perspectives. We use experiments on real-world merchant transaction data to demonstrate the effectiveness of our proposed model. We also provide extensive discussions on different model design choices in our experimental section.
Audio captioning is a multi-modal task, focusing on using natural language for describing the contents of general audio. Most audio captioning methods are based on deep neural networks, employing an encoder-decoder scheme and a dataset with audio clips and corresponding natural language descriptions (i.e. captions). A significant challenge for audio captioning is the distribution of words in the captions: some words are very frequent but acoustically non-informative, i.e. the function words (e.g. "a", "the"), and other words are infrequent but informative, i.e. the content words (e.g. adjectives, nouns). In this paper we propose two methods to mitigate this class imbalance problem. First, in an autoencoder setting for audio captioning, we weigh each word's contribution to the training loss inversely proportional to its number of occurrences in the whole dataset. Secondly, in addition to multi-class, word-level audio captioning task, we define a multi-label side task based on clip-level content word detection by training a separate decoder. We use the loss from the second task to regularize the jointly trained encoder for the audio captioning task. We evaluate our method using Clotho, a recently published, wide-scale audio captioning dataset, and our results show an increase of 37\% relative improvement with SPIDEr metric over the baseline method.
Domestic service robots (DSRs) are a promising solution to the shortage of home care workers. However, one of the main limitations of DSRs is their inability to interact naturally through language. Recently, data-driven approaches have been shown to be effective for tackling this limitation; however, they often require large-scale datasets, which is costly. Based on this background, we aim to perform automatic sentence generation of fetching instructions: for example, "Bring me a green tea bottle on the table." This is particularly challenging because appropriate expressions depend on the target object, as well as its surroundings. In this paper, we propose the attention branch encoder--decoder network (ABEN), to generate sentences from visual inputs. Unlike other approaches, the ABEN has multimodal attention branches that use subword-level attention and generate sentences based on subword embeddings. In experiments, we compared the ABEN with a baseline method using four standard metrics in image captioning. Results show that the ABEN outperformed the baseline in terms of these metrics.
Given a speaker's speech, it is interesting to see if it is possible to generate this speaker's face. One main challenge in this task is to alleviate the natural mismatch between face and speech. To this end, in this paper, we propose a novel Attention-based Residual Speech Portrait Model (AR-SPM) by introducing the ideal of the residual into a hybrid encoder-decoder architecture, where face prior features are merged with the output of speech encoder to form the final face feature. In particular, we innovatively establish a tri-item loss function, which is a weighted linear combination of the L2-norm, L1-norm and negative cosine loss, to train our model by comparing the final face feature and true face feature. Evaluation on AVSpeech dataset shows that our proposed model accelerates the convergence of training, outperforms the state-of-the-art in terms of quality of the generated face, and achieves superior recognition accuracy of gender and age compared with the ground truth.
Different from image inpainting, image outpainting has relative less context in the image center to capture and more content at the image border to predict. Therefore, classical encoder-decoder pipeline of existing methods may not predict the outstretched unknown content perfectly. In this paper, a novel two-stage siamese adversarial model for image extrapolation, named Siamese Expansion Network (SiENet) is proposed. In two stages, a novel border sensitive convolution named adaptive filling convolution is designed for allowing encoder to predict the unknown content, alleviating the burden of decoder. Besides, to introduce prior knowledge to network and reinforce the inferring ability of encoder, siamese adversarial mechanism is designed to enable our network to model the distribution of covered long range feature for that of uncovered image feature. The results on four datasets has demonstrated that our method outperforms existing state-of-the-arts and could produce realistic results.
Decompressive craniectomy (DC) is a common surgical procedure consisting of the removal of a portion of the skull that is performed after incidents such as stroke, traumatic brain injury (TBI) or other events that could result in acute subdural hemorrhage and/or increasing intracranial pressure. In these cases, CT scans are obtained to diagnose and assess injuries, or guide a certain therapy and intervention. We propose a deep learning based method to reconstruct the skull defect removed during DC performed after TBI from post-operative CT images. This reconstruction is useful in multiple scenarios, e.g. to support the creation of cranioplasty plates, accurate measurements of bone flap volume and total intracranial volume, important for studies that aim to relate later atrophy to patient outcome. We propose and compare alternative self-supervised methods where an encoder-decoder convolutional neural network (CNN) estimates the missing bone flap on post-operative CTs. The self-supervised learning strategy only requires images with complete skulls and avoids the need for annotated DC images. For evaluation, we employ real and simulated images with DC, comparing the results with other state-of-the-art approaches. The experiments show that the proposed model outperforms current manual methods, enabling reconstruction even in highly challenging cases where big skull defects have been removed during surgery.
Since the mid-2000s, there has been a resurrection of interest in modern modifications of fiducial inference. To date, the main computational tool to extract a generalized fiducial distribution is Markov chain Monte Carlo (MCMC). We propose an alternative way of computing a generalized fiducial distribution that could be used in complex situations. In particular, to overcome the difficulty when the unnormalized fiducial density (needed for MCMC), we design a fiducial autoencoder (FAE). The fitted autoencoder is used to generate generalized fiducial samples of the unknown parameters. To increase accuracy, we then apply an approximate fiducial computation (AFC) algorithm, by rejecting samples that when plugged into a decoder do not replicate the observed data well enough. Our numerical experiments show the effectiveness of our FAE-based inverse solution and the excellent coverage performance of the AFC corrected FAE solution.
We propose a novel neural network architecture, called autoencoder-constrained graph convolutional network, to solve node classification task on graph domains. As suggested by its name, the core of this model is a convolutional network operating directly on graphs, whose hidden layers are constrained by an autoencoder. Comparing with vanilla graph convolutional networks, the autoencoder step is added to reduce the information loss brought by Laplacian smoothing. We consider applying our model on both homogeneous graphs and heterogeneous graphs. For homogeneous graphs, the autoencoder approximates to the adjacency matrix of the input graph by taking hidden layer representations as encoder and another one-layer graph convolutional network as decoder. For heterogeneous graphs, since there are multiple adjacency matrices corresponding to different types of edges, the autoencoder approximates to the feature matrix of the input graph instead, and changes the encoder to a particularly designed multi-channel pre-processing network with two layers. In both cases, the error occurred in the autoencoder approximation goes to the penalty term in the loss function. In extensive experiments on citation networks and other heterogeneous graphs, we demonstrate that adding autoencoder constraints significantly improves the performance of graph convolutional networks. Further, we notice that our technique can be applied on graph attention network to improve the performance as well. This reveals the wide applicability of the proposed autoencoder technique.
Modeling dynamics of human motion is one of the most challenging sequence modeling problem, with diverse applications in animation industry, human-robot interaction, motion-based surveillance, etc. Available attempts to use auto-regressive techniques for long-term single-person motion generation usually fails, resulting in stagnated motion or divergence to unrealistic pose patterns. In this paper, we propose a novel cross-conditioned recurrent framework targeting long-term synthesis of inter-person interactions beyond several minutes. We carefully integrate positive implications of both auto-regressive and encoder-decoder recurrent architecture, by interchangeably utilizing two separate fixed-length cross person motion prediction models for long-term generation in a novel hierarchical fashion. As opposed to prior approaches, we guarantee structural plausibility of 3D pose by training the recurrent model to regress latent representation of a separately trained generative pose embedding network. Different variants of the proposed frameworks are evaluated through extensive experiments on SBU-interaction, CMU-MoCAP and an inhouse collection of duet-dance dataset. Qualitative and quantitative evaluation on several tasks, such as Short-term motion prediction, Long-term motion synthesis and Interaction-based motion retrieval against prior state-of-the-art approaches clearly highlight superiority of the proposed framework.
Printed Mathematical expression recognition (PMER) aims to transcribe a printed mathematical expression image into a structural expression, such as LaTeX expression. It is a crucial task for many applications, including automatic question recommendation, automatic problem solving and analysis of the students, etc. Currently, the mainstream solutions rely on solving image captioning tasks, all addressing image summarization. As such, these methods can be suboptimal for solving MER problem. In this paper, we propose a new method named EDSL, shorted for encoder-decoder with symbol-level features, to identify the printed mathematical expressions from images. The symbol-level image encoder of EDSL consists of segmentation module and reconstruction module. By performing segmentation module, we identify all the symbols and their spatial information from images in an unsupervised manner. We then design a novel reconstruction module to recover the symbol dependencies after symbol segmentation. Especially, we employ a position correction attention mechanism to capture the spatial relationships between symbols. To alleviate the negative impact from long output, we apply the transformer model for transcribing the encoded image into the sequential and structural output. We conduct extensive experiments on two real datasets to verify the effectiveness and rationality of our proposed EDSL method. The experimental results have illustrated that EDSL has achieved 92.7\% and 89.0\% in evaluation metric Match, which are 3.47\% and 4.04\% higher than the state-of-the-art method. Our code and datasets are available at https://github.com/abcAnonymous/EDSL .
Tools capable of automatic code generation have the potential to augment programmer's capabilities. While straightforward code retrieval is incorporated into many IDEs, an emerging area is explicit code generation. Code generation is currently approached as a Machine Translation task, with Recurrent Neural Network (RNN) based encoder-decoder architectures trained on code-description pairs. In this work we introduce and study modern Transformer architectures for this task. We further propose a new model called the Relevance Transformer that incorporates external knowledge using pseudo-relevance feedback. The Relevance Transformer biases the decoding process to be similar to existing retrieved code while enforcing diversity. We perform experiments on multiple standard benchmark datasets for code generation including Django, Hearthstone, and CoNaLa. The results show improvements over state-of-the-art methods based on BLEU evaluation. The Relevance Transformer model shows the potential of Transformer-based architectures for code generation and introduces a method of incorporating pseudo-relevance feedback during inference.
Convolutional Neural Networks (CNNs) are highly effective for image reconstruction problems. Typically, CNNs are trained on large amounts of training images. Recently, however, un-trained CNNs such as the Deep Image Prior and Deep Decoder have achieved excellent performance for image reconstruction problems such as denoising and inpainting, \emph{without using any training data}. Motivated by this development, we address the reconstruction problem arising in accelerated MRI with un-trained neural networks. We propose a highly optimized un-trained recovery approach based on a variation of the Deep Decoder and show that it significantly outperforms other un-trained methods, in particular sparsity-based classical compressed sensing methods and naive applications of un-trained neural networks. We also compare performance (both in terms of reconstruction accuracy and computational cost) in an ideal setup for trained methods, specifically on the fastMRI dataset, where the training and test data come from the same distribution. We find that our un-trained algorithm achieves similar performance to a baseline trained neural network, but a state-of-the-art trained network outperforms the un-trained one. Finally, we perform a comparison on a non-ideal setup where the train and test distributions are slightly different, and find that our un-trained method achieves similar performance to a state-of-the-art accelerated MRI reconstruction method.
In this work, we present Auto-captions on GIF, which is a new large-scale pre-training dataset for generic video understanding. All video-sentence pairs are created by automatically extracting and filtering video caption annotations from billions of web pages. Auto-captions on GIF dataset can be utilized to pre-train the generic feature representation or encoder-decoder structure for video captioning, and other downstream tasks (e.g., sentence localization in videos, video question answering, etc.) as well. We present a detailed analysis of Auto-captions on GIF dataset in comparison to existing video-sentence datasets. We also provide an evaluation of a Transformer-based encoder-decoder structure for vision-language pre-training, which is further adapted to video captioning downstream task and yields the compelling generalizability on MSR-VTT. The dataset is available at \url{http://www.auto-video-captions.top/2020/dataset}.