Autoencoders are data-specific compression algorithms learned automatically from examples. The predominant approach has been to construct single large global models that cover the domain. However, training and evaluating models of increasing size comes at the price of additional time and computational cost. Conditional computation, sparsity, and model pruning techniques can reduce these costs while maintaining performance. Learning classifier systems (LCS) are a framework for adaptively subdividing input spaces into an ensemble of simpler local approximations that together cover the domain. LCS perform conditional computation through the use of a population of individual gating/guarding components, each associated with a local approximation. This article explores the use of an LCS to adaptively decompose the input domain into a collection of small autoencoders where local solutions of different complexity may emerge. In addition to benefits in convergence time and computational cost, it is shown possible to reduce code size as well as the resulting decoder computational cost when compared with the global model equivalent.
When a channel model is available, learning how to communicate on fading noisy channels can be formulated as the (unsupervised) training of an autoencoder consisting of the cascade of encoder, channel, and decoder. An important limitation of the approach is that training should be generally carried out from scratch for each new channel. To cope with this problem, prior works considered joint training over multiple channels with the aim of finding a single pair of encoder and decoder that works well on a class of channels. As a result, joint training ideally mimics the operation of non-coherent transmission schemes. In this paper, we propose to obviate the limitations of joint training via meta-learning: Rather than training a common model for all channels, meta-learning finds a common initialization vector that enables fast training on any channel. The approach is validated via numerical results, demonstrating significant training speed-ups, with effective encoders and decoders obtained with as little as one iteration of Stochastic Gradient Descent.
With the rapid development of mobile-internet technologies, on-demand ride-sourcing services have become increasingly popular and largely reshaped the way people travel. Demand prediction is one of the most fundamental components in supply-demand management systems of ride-sourcing platforms. With accurate short-term prediction for origin-destination (OD) demand, the platforms make precise and timely decisions on real-time matching, idle vehicle reallocations and ride-sharing vehicle routing, etc. Compared to zone-based demand prediction that has been examined by many previous studies, OD-based demand prediction is more challenging. This is mainly due to the complicated spatial and temporal dependencies among demand of different OD pairs. To overcome this challenge, we propose the Spatio-Temporal Encoder-Decoder Residual Multi-Graph Convolutional network (ST-ED-RMGC), a novel deep learning model for predicting ride-sourcing demand of various OD pairs. Firstly, the model constructs OD graphs, which utilize adjacent matrices to characterize the non-Euclidean pair-wise geographical and semantic correlations among different OD pairs. Secondly, based on the constructed graphs, a residual multi-graph convolutional (RMGC) network is designed to encode the contextual-aware spatial dependencies, and a long-short term memory (LSTM) network is used to encode the temporal dependencies, into a dense vector space. Finally, we reuse the RMGC networks to decode the compressed vector back to OD graphs and predict the future OD demand. Through extensive experiments on the for-hire-vehicles datasets in Manhattan, New York City, we show that our proposed deep learning framework outperforms the state-of-arts by a significant margin.
Clustering high-dimensional data, such as images or biological measurements, is a long-standingproblem and has been studied extensively. Recently, Deep Clustering has gained popularity due toits flexibility in fitting the specific peculiarities of complex data. Here we introduce the Mixture-of-Experts Similarity Variational Autoencoder (MoE-Sim-VAE), a novel generative clustering model.The model can learn multi-modal distributions of high-dimensional data and use these to generaterealistic data with high efficacy and efficiency. MoE-Sim-VAE is based on a Variational Autoencoder(VAE), where the decoder consists of a Mixture-of-Experts (MoE) architecture. This specific architecture allows for various modes of the data to be automatically learned by means of the experts.Additionally, we encourage the lower dimensional latent representation of our model to follow aGaussian mixture distribution and to accurately represent the similarities between the data points. Weassess the performance of our model on the MNIST benchmark data set and challenging real-worldtasks of clustering mouse organs from single-cell RNA-sequencing measurements and defining cellsubpopulations from mass cytometry (CyTOF) measurements on hundreds of different datasets.MoE-Sim-VAE exhibits superior clustering performance on all these tasks in comparison to thebaselines as well as competitor methods.
Recently, deep neural networks have demonstrated comparable and even better performance with board-certified ophthalmologists in well-annotated datasets. However, the diversity of retinal imaging devices poses a significant challenge: domain shift, which leads to performance degradation when applying the deep learning models to new testing domains. In this paper, we propose a novel unsupervised domain adaptation framework, called Collaborative Feature Ensembling Adaptation (CFEA), to effectively overcome this challenge. Our proposed CFEA is an interactive paradigm which presents an exquisite of collaborative adaptation through both adversarial learning and ensembling weights. In particular, we simultaneously achieve domain-invariance and maintain an exponential moving average of the historical predictions, which achieves a better prediction for the unlabeled data, via ensembling weights during training. Without annotating any sample from the target domain, multiple adversarial losses in encoder and decoder layers guide the extraction of domain-invariant features to confuse the domain classifier and meanwhile benefit the ensembling of smoothing weights. Comprehensive experimental results demonstrate that our CFEA model can overcome performance degradation and outperform the state-of-the-art methods in segmenting retinal optic disc and cup from fundus images. \textit{Code is available at \url{https://github.com/cswin/AWC}}.
We evaluate three simple, normalization-centric changes to improve Transformer training. First, we show that pre-norm residual connections (PreNorm) and smaller initializations enable warmup-free, validation-based training with large learning rates. Second, we propose $\ell_2$ normalization with a single scale parameter (ScaleNorm) for faster training and better performance. Finally, we reaffirm the effectiveness of normalizing word embeddings to a fixed length (FixNorm). On five low-resource translation pairs from TED Talks-based corpora, these changes always converge, giving an average +1.1 BLEU over state-of-the-art bilingual baselines and a new 32.8 BLEU on IWSLT'15 English-Vietnamese. We observe sharper performance curves, more consistent gradient norms, and a linear relationship between activation scaling and decoder depth. Surprisingly, in the high-resource setting (WMT'14 English-German), ScaleNorm and FixNorm remain competitive but PreNorm degrades performance.
We present Panoptic-DeepLab, a bottom-up and single-shot approach for panoptic segmentation. Our Panoptic-DeepLab is conceptually simple and delivers state-of-the-art results. In particular, we adopt the dual-ASPP and dual-decoder structures specific to semantic, and instance segmentation, respectively. The semantic segmentation branch is the same as the typical design of any semantic segmentation model (e.g., DeepLab), while the instance segmentation branch is class-agnostic, involving a simple instance center regression. Our single Panoptic-DeepLab sets the new state-of-art at all three Cityscapes benchmarks, reaching 84.2% mIoU, 39.0% AP, and 65.5% PQ on test set, and advances results on the other challenging Mapillary Vistas.
We introduce the Mutual Information Machine (MIM), a probabilistic auto-encoder for learning joint distributions over observations and latent variables. MIM reflects three design principles: 1) low divergence, to encourage the encoder and decoder to learn consistent factorizations of the same underlying distribution; 2) high mutual information, to encourage an informative relation between data and latent variables; and 3) low marginal entropy, or compression, which tends to encourage clustered latent representations. We show that a combination of the Jensen-Shannon divergence and the joint entropy of the encoding and decoding distributions satisfies these criteria, and admits a tractable cross-entropy bound that can be optimized directly with Monte Carlo and stochastic gradient descent. We contrast MIM learning with maximum likelihood and VAEs. Experiments show that MIM learns representations with high mutual information, consistent encoding and decoding distributions, effective latent clustering, and data log likelihood comparable to VAE, while avoiding posterior collapse.
This paper studies the classification problem on electroencephalogram (EEG) data of mental tasks, using standard architecture of three-layer CNN, stacked LSTM, stacked GRU. We further propose a novel classifier - a mixed LSTM model with a CNN decoder. A hyperparameter optimization on CNN shows validation accuracy of 72% and testing accuracy of 62%. The stacked LSTM and GRU models with FFT preprocessing and downsampling on data achieve 55% and 51% testing accuracy respectively. As for the mixed LSTM model with CNN decoder, validation accuracy of 75% and testing accuracy of 70% are obtained. We believe the mixed model is more robust and accurate than both CNN and LSTM individually, by using the CNN layer as a decoder for following LSTM layers. The code is completed in the framework of Pytorch and Keras. Results and code can be found at https://github.com/theyou21/BigProject.
Attention-based scene text recognizers have gained huge success, which leverages a more compact intermediate representation to learn 1d- or 2d- attention by a RNN-based encoder-decoder architecture. However, such methods suffer from attention-drift problem because high similarity among encoded features leads to attention confusion under the RNN-based local attention mechanism. Moreover, RNN-based methods have low efficiency due to poor parallelization. To overcome these problems, we propose the MASTER, a self-attention based scene text recognizer that (1) not only encodes the input-output attention but also learns self-attention which encodes feature-feature and target-target relationships inside the encoder and decoder and (2) learns a more powerful and robust intermediate representation to spatial distortion, and (3) owns a great training efficiency because of high training parallelization and a high-speed inference because of an efficient memory-cache mechanism. Extensive experiments on various benchmarks demonstrate the superior performance of our MASTER on both regular and irregular scene text. Pytorch code can be found at https://github.com/wenwenyu/MASTER-pytorch, and Tensorflow code can be found at https://github.com/jiangxiluning/MASTER-TF.
While generative models have shown great success in generating high-dimensional samples conditional on low-dimensional descriptors (learning e.g. stroke thickness in MNIST, hair color in CelebA, or speaker identity in Wavenet), their generation out-of-sample poses fundamental problems. The conditional variational autoencoder (CVAE) as a simple conditional generative model does not explicitly relate conditions during training and, hence, has no incentive of learning a compact joint distribution across conditions. We overcome this limitation by matching their distributions using maximum mean discrepancy (MMD) in the decoder layer that follows the bottleneck. This introduces a strong regularization both for reconstructing samples within the same condition and for transforming samples across conditions, resulting in much improved generalization. We refer to the architecture as \emph{transformer} VAE (trVAE). Benchmarking trVAE on high-dimensional image and tabular data, we demonstrate higher robustness and higher accuracy than existing approaches. In particular, we show qualitatively improved predictions for cellular perturbation response to treatment and disease based on high-dimensional single-cell gene expression data, by tackling previously problematic minority classes and multiple conditions. For generic tasks, we improve Pearson correlations of high-dimensional estimated means and variances with their ground truths from 0.89 to 0.97 and 0.75 to 0.87, respectively.
Simultaneous machine translation models start generating a target sequence before they have encoded or read the source sequence. Recent approaches for this task either apply a fixed policy on a state-of-the art Transformer model, or a learnable monotonic attention on a weaker recurrent neural network-based structure. In this paper, we propose a new attention mechanism, Monotonic Multihead Attention (MMA), which extends the monotonic attention mechanism to multihead attention. We also introduce two novel and interpretable approaches for latency control that are specifically designed for multiple attentions heads. We apply MMA to the simultaneous machine translation task and demonstrate better latency-quality tradeoffs compared to MILk, the previous state-of-the-art approach. We also analyze how the latency controls affect the attention span and we motivate the introduction of our model by analyzing the effect of the number of decoder layers and heads on quality and latency.
We study open domain response generation with limited message-response pairs. The problem exists in real-world applications but is less explored by the existing work. Since the paired data now is no longer enough to train a neural generation model, we consider leveraging the large scale of unpaired data that are much easier to obtain, and propose response generation with both paired and unpaired data. The generation model is defined by an encoder-decoder architecture with templates as prior, where the templates are estimated from the unpaired data as a neural hidden semi-markov model. By this means, response generation learned from the small paired data can be aided by the semantic and syntactic knowledge in the large unpaired data. To balance the effect of the prior and the input message to response generation, we propose learning the whole generation model with an adversarial approach. Empirical studies on question response generation and sentiment response generation indicate that when only a few pairs are available, our model can significantly outperform several state-of-the-art response generation models in terms of both automatic and human evaluation.
We propose a Dual-Stream Pyramid Registration Network (referred as Dual-PRNet) for unsupervised 3D medical image registration. Unlike recent CNN-based registration approaches, such as VoxelMorph, which explores a single-stream encoder-decoder network to compute a registration fields from a pair of 3D volumes, we design a two-stream architecture able to compute multi-scale registration fields from convolutional feature pyramids. Our contributions are two-fold: (i) we design a two-stream 3D encoder-decoder network which computes two convolutional feature pyramids separately for a pair of input volumes, resulting in strong deep representations that are meaningful for deformation estimation; (ii) we propose a pyramid registration module able to predict multi-scale registration fields directly from the decoding feature pyramids. This allows it to refine the registration fields gradually in a coarse-to-fine manner via sequential warping, and enable the model with the capability for handling significant deformations between two volumes, such as large displacements in spatial domain or slice space. The proposed Dual-PRNet is evaluated on two standard benchmarks for brain MRI registration, where it outperforms the state-of-the-art approaches by a large margin, e.g., having improvements over recent VoxelMorph [2] with 0.683->0.778 on the LPBA40, and 0.511->0.631 on the Mindboggle101, in term of average Dice score. Code is available at: https://github.com/kangmiao15/Dual-Stream-PRNet-Plus.
This paper introduces the problem of multiple object forecasting (MOF), in which the goal is to predict future bounding boxes of tracked objects. In contrast to existing works on object trajectory forecasting which primarily consider the problem from a birds-eye perspective, we formulate the problem from an object-level perspective and call for the prediction of full object bounding boxes, rather than trajectories alone. Towards solving this task, we introduce the Citywalks dataset, which consists of over 200k high-resolution video frames. Citywalks comprises of footage recorded in 21 cities from 10 European countries in a variety of weather conditions and over 3.5k unique pedestrian trajectories. For evaluation, we adapt existing trajectory forecasting methods for MOF and confirm cross-dataset generalizability on the MOT-17 dataset without fine-tuning. Finally, we present STED, a novel encoder-decoder architecture for MOF. STED combines visual and temporal features to model both object-motion and ego-motion, and outperforms existing approaches for MOF. Code & dataset link: https://github.com/olly-styles/Multiple-Object-Forecasting
In this paper, we address the semantic segmentation problem with a focus on the context aggregation strategy. Our motivation is that the label of a pixel is the category of the object that the pixel belongs to. We present a simple yet effective approach, object-contextual representations, characterizing a pixel by exploiting the representation of the corresponding object class. First, we learn object regions under the supervision of ground-truth segmentation. Second, we compute the object region representation by aggregating the representations of the pixels lying in the object region. Last, % the representation similarity we compute the relation between each pixel and each object region and augment the representation of each pixel with the object-contextual representation which is a weighted aggregation of all the object region representations according to their relations with the pixel. We empirically demonstrate that the proposed approach achieves competitive performance on various challenging semantic segmentation benchmarks: Cityscapes, ADE20K, LIP, PASCAL-Context, and COCO-Stuff. Cityscapes, ADE20K, LIP, PASCAL-Context, and COCO-Stuff. Our submission "HRNet + OCR + SegFix" achieves 1-st place on the Cityscapes leaderboard by the time of submission. Code is available at: https://git.io/openseg and https://git.io/HRNet.OCR. We rephrase the object-contextual representation scheme using the Transformer encoder-decoder framework. The details are presented in~Section3.3.
This paper presents a unified Vision-Language Pre-training (VLP) model. The model is unified in that (1) it can be fine-tuned for either vision-language generation (e.g., image captioning) or understanding (e.g., visual question answering) tasks, and (2) it uses a shared multi-layer transformer network for both encoding and decoding, which differs from many existing methods where the encoder and decoder are implemented using separate models. The unified VLP model is pre-trained on a large amount of image-text pairs using the unsupervised learning objectives of two tasks: bidirectional and sequence-to-sequence (seq2seq) masked vision-language prediction. The two tasks differ solely in what context the prediction conditions on. This is controlled by utilizing specific self-attention masks for the shared transformer network. To the best of our knowledge, VLP is the first reported model that achieves state-of-the-art results on both vision-language generation and understanding tasks, as disparate as image captioning and visual question answering, across three challenging benchmark datasets: COCO Captions, Flickr30k Captions, and VQA 2.0. The code and the pre-trained models are available at https://github.com/LuoweiZhou/VLP.
In this work we focus on transferring supervision signals of natural language generation (NLG) tasks between multiple languages. We propose to pretrain the encoder and the decoder of a sequence-to-sequence model under both monolingual and cross-lingual settings. The pre-training objective encourages the model to represent different languages in the shared space, so that we can conduct zero-shot cross-lingual transfer. After the pre-training procedure, we use monolingual data to fine-tune the pre-trained model on downstream NLG tasks. Then the sequence-to-sequence model trained in a single language can be directly evaluated beyond that language (i.e., accepting multi-lingual input and producing multi-lingual output). Experimental results on question generation and abstractive summarization show that our model outperforms the machine-translation-based pipeline methods for zero-shot cross-lingual generation. Moreover, cross-lingual transfer improves NLG performance of low-resource languages by leveraging rich-resource language data. Our implementation and data are available at https://github.com/CZWin32768/xnlg.
In this paper, we focus on devising a versatile framework for dense pixelwise prediction whose goal is to assign a discrete or continuous label to each pixel for an image. It is well-known that the reduced feature resolution due to repeated subsampling operations poses a serious challenge to Fully Convolutional Network (FCN) based models. In contrast to the commonly-used strategies, such as dilated convolution and encoder-decoder structure, we introduce the Flattening Module to produce high-resolution predictions without either removing any subsampling operations or building a complicated decoder module. In addition, the Flattening Module is lightweight and can be easily combined with any existing FCNs, allowing the model builder to trade off among model size, computational cost and accuracy by simply choosing different backbone networks. We empirically demonstrate the effectiveness of the proposed Flattening Module through competitive results in human pose estimation on MPII, semantic segmentation on PASCAL-Context and object detection on PASCAL VOC. We hope that the proposed approach can serve as a simple and strong alternative of current dominant dense pixelwise prediction frameworks.
Natural image matting is an important problem in computer vision and graphics. It is an ill-posed problem when only an input image is available without any external information. While the recent deep learning approaches have shown promising results, they only estimate the alpha matte. This paper presents a context-aware natural image matting method for simultaneous foreground and alpha matte estimation. Our method employs two encoder networks to extract essential information for matting. Particularly, we use a matting encoder to learn local features and a context encoder to obtain more global context information. We concatenate the outputs from these two encoders and feed them into decoder networks to simultaneously estimate the foreground and alpha matte. To train this whole deep neural network, we employ both the standard Laplacian loss and the feature loss: the former helps to achieve high numerical performance while the latter leads to more perceptually plausible results. We also report several data augmentation strategies that greatly improve the network's generalization performance. Our qualitative and quantitative experiments show that our method enables high-quality matting for a single natural image. Our inference codes and models have been made publicly available at https://github.com/hqqxyy/Context-Aware-Matting.