Object detection in remote sensing, especially in aerial images, remains a challenging problem due to low image resolution, complex backgrounds, and variation of scale and angles of objects in images. In current implementations, multi-scale based and angle-based networks have been proposed and generate promising results with aerial image detection. In this paper, we propose a novel loss function, called Salience Biased Loss (SBL), for deep neural networks, which uses salience information of the input image to achieve improved performance for object detection. Our novel loss function treats training examples differently based on input complexity in order to avoid the over-contribution of easy cases in the training process. In our experiments, RetinaNet was trained with SBL to generate an one-stage detector, SBL-RetinaNet. SBL-RetinaNet is applied to the largest existing public aerial image dataset, DOTA. Experimental results show our proposed loss function with the RetinaNet architecture outperformed other state-of-art object detection models by at least 4.31 mAP, and RetinaNet by 2.26 mAP with the same inference speed of RetinaNet.
https://arxiv.org/abs/1810.08103
Among underwater perceptual sensors, imaging sonar has been highlighted for its perceptual robustness underwater. The major challenge of imaging sonar, however, arises from the difficulty in defining visual features despite limited resolution and high noise levels. Recent developments in deep learning provide a powerful solution for computer-vision researches using optical images. Unfortunately, deep learning-based approaches are not well established for imaging sonars, mainly due to the scant data in the training phase. Unlike the abundant publically available terrestrial images, obtaining underwater images is often costly, and securing enough underwater images for training is not straightforward. To tackle this issue, this paper presents a solution to this field’s lack of data by introducing a novel end-to-end image-synthesizing method in the training image preparation phase. The proposed method present image synthesizing scheme to the images captured by an underwater simulator. Our synthetic images are based on the sonar imaging models and noisy characteristics to represent the real data obtained from the sea. We validate the proposed scheme by training using a simulator and by testing the simulated images with real underwater sonar images obtained from a water tank and the sea.
https://arxiv.org/abs/1810.07990
Accurate and fast extraction of the foreground object is one of the most significant issues to be solved due to its important meaning for object tracking and recognition in video surveillance. Although many foreground object detection methods have been proposed in the recent past, it is still regarded as a tough problem due to illumination variations and dynamic backgrounds challenges. In this paper, we propose a robust foreground object detection method with two aspects of contributions. First, we propose a robust texture operator named Robust Local Binary Similarity Pattern (RLBSP), which shows strong robustness to illumination variations and dynamic backgrounds. Second, a combination of color and texture features are used to characterize pixel representations, which compensate each other to make full use of their own advantages. Comprehensive experiments evaluated on the CDnet 2012 dataset demonstrate that the proposed method performs favorably against state-of-the-art methods.
https://arxiv.org/abs/1810.06797
A natural way to improve the detection of objects is to consider the contextual constraints imposed by the detection of additional objects in a given scene. In this work, we exploit the spatial relations between objects in order to improve detection capacity, as well as analyze various properties of the contextual object detection problem. To precisely calculate context-based probabilities of objects, we developed a model that examines the interactions between objects in an exact probabilistic setting, in contrast to previous methods that typically utilize approximations based on pairwise interactions. Such a scheme is facilitated by the realistic assumption that the existence of an object in any given location is influenced by only few informative locations in space. Based on this assumption, we suggest a method for identifying these relevant locations and integrating them into a mostly exact calculation of probability based on their raw detector responses. This scheme is shown to improve detection results and provides unique insights about the process of contextual inference for object detection. We show that it is generally difficult to learn that a particular object reduces the probability of another, and that in cases when the context and detector strongly disagree this learning becomes virtually impossible for the purposes of improving the results of an object detector. Finally, we demonstrate improved detection results through use of our approach as applied to the PASCAL VOC and COCO datasets.
https://arxiv.org/abs/1711.05705
Recent work has shown that the encoder-decoder attention mechanisms in neural machine translation (NMT) are different from the word alignment in statistical machine translation. In this paper, we focus on analyzing encoder-decoder attention mechanisms, in the case of word sense disambiguation (WSD) in NMT models. We hypothesize that attention mechanisms pay more attention to context tokens when translating ambiguous words. We explore the attention distribution patterns when translating ambiguous nouns. Counter-intuitively, we find that attention mechanisms are likely to distribute more attention to the ambiguous noun itself rather than context tokens, in comparison to other nouns. We conclude that attention mechanism is not the main mechanism used by NMT models to incorporate contextual information for WSD. The experimental results suggest that NMT models learn to encode contextual information necessary for WSD in the encoder hidden states. For the attention mechanism in Transformer models, we reveal that the first few layers gradually learn to “align” source and target tokens and the last few layers learn to extract features from the related but unaligned context tokens.
https://arxiv.org/abs/1810.07595
Most investigations into near-memory hardware accelerators for deep neural networks have primarily focused on inference, while the potential of accelerating training has received relatively little attention so far. Based on an in-depth analysis of the key computational patterns in state-of-the-art gradient-based training methods, we propose an efficient near-memory acceleration engine called NTX that can be used to train state-of-the-art deep convolutional neural networks at scale. Our main contributions are: (i) a loose coupling of RISC-V cores and NTX co-processors reducing offloading overhead by 7x over previously published results; (ii) an optimized IEEE754 compliant data path for fast high-precision convolutions and gradient propagation; (iii) evaluation of near-memory computing with NTX embedded into residual area on the Logic Base die of a Hybrid Memory Cube; and (iv) a scaling analysis to meshes of HMCs in a data center scenario. We demonstrate a 2.7x energy efficiency improvement of NTX over contemporary GPUs at 4.4x less silicon area, and a compute performance of 1.2 Tflop/s for training large state-of-the-art networks with full floating-point precision. At the data center scale, a mesh of NTX achieves above 95% parallel and energy efficiency, while providing 2.1x energy savings or 3.1x performance improvement over a GPU-based system.
https://arxiv.org/abs/1803.04783
We establish a general linear response relation for spiking neuronal networks, based on chains with unbounded memory. This relation allows quantifying the influence of a weak amplitude external stimuli on spatio-temporal spike correlations, in a general context where the memory in spike dynamics can go arbitrarily far in the past. With this approach, we show how linear response is explicitly related to neuron dynamics with an example, the gIF model, introduced by M. Rudolph and A. Destexhe [91]. This illustrates the effect of the stimuli, intrinsic neuronal dynamics, and network connectivity on spike statistics.
https://arxiv.org/abs/1704.05344
Visual data and text data are composed of information at multiple granularities. A video can describe a complex scene that is composed of multiple clips or shots, where each depicts a semantically coherent event or action. Similarly, a paragraph may contain sentences with different topics, which collectively conveys a coherent message or story. In this paper, we investigate the modeling techniques for such hierarchical sequential data where there are correspondences across multiple modalities. Specifically, we introduce hierarchical sequence embedding (HSE), a generic model for embedding sequential data of different modalities into hierarchically semantic spaces, with either explicit or implicit correspondence information. We perform empirical studies on large-scale video and paragraph retrieval datasets and demonstrated superior performance by the proposed methods. Furthermore, we examine the effectiveness of our learned embeddings when applied to downstream tasks. We show its utility in zero-shot action recognition and video captioning.
https://arxiv.org/abs/1810.07212
Detection of salient objects in image and video is of great importance in many computer vision applications. In spite of the fact that the state of the art in saliency detection for still images has been changed substantially over the last few years, there have been few improvements in video saliency detection. This paper investigates the use of recently introduced non-local neural networks in video salient object detection. Non-local neural networks are applied to capture global dependencies and hence determine the salient objects. The effect of non-local operations is studied separately on static and dynamic saliency detection in order to exploit both appearance and motion features. A novel deep non-local neural network architecture is introduced for video salient object detection and tested on two well-known datasets DAVIS and FBMS. The experimental results show that the proposed algorithm outperforms state-of-the-art video saliency detection methods.
https://arxiv.org/abs/1810.07097
We incorporate an explicit neural interlingua into a multilingual encoder-decoder neural machine translation (NMT) architecture. We demonstrate that our model learns a language-independent representation by performing direct zero-shot translation (without using pivot translation), and by using the source sentence embeddings to create an English Yelp review classifier that, through the mediation of the neural interlingua, can also classify French and German reviews. Furthermore, we show that, despite using a smaller number of parameters than a pairwise collection of bilingual NMT models, our approach produces comparable BLEU scores for each language pair in WMT15.
https://arxiv.org/abs/1804.08198
High quality semipolar and nonpolar GaN is crucial in achieving high-performance GaN-based optoelectronic devices, yet it has been very challenging to achieve large-area wafers that are free of basal-plane stacking faults (BSFs). In this work, we report an approach to prepare large-area, stacking-fault-free (SF-free) semipolar GaN on (4-inch) sapphire substrates. A root cause of the formation of BSFs is the emergence of N-polar (000-1) facets during semipolar and non-polar heteroepitaxy. Invoking the concept of kinetic Wulff plot, we succeeded in suppressing the occurrence of N-polar GaN (000-1) facets, and consequently in eliminating the stacking faults generated in (000-1) basal-planes. The result was confirmed by transmission electron microscopy, cathodoluminescence, and low-temperature photoluminescence characterizations. Furthermore, InGaN light emitting diodes with promising characteristics have been produced on the SF-free semipolar (20-21) GaN on sapphire substrates. Our work opens up a new insight about the heteroepitaxial growth of nonpolar/semipolar GaN and provides an approach of producing SF-free nonpolar/semipolar GaN material over large-area wafers which will create new opportunities in GaN optoelectronic and microelectronic research.
https://arxiv.org/abs/1810.06794
Neural machine translation (NMT) is notoriously sensitive to noises, but noises are almost inevitable in practice. One special kind of noise is the homophone noise, where words are replaced by other words with the same (or similar) pronunciations. Homophone noise arises frequently from many real-world scenarios upstream to translation, such as automatic speech recognition (ASR) or phonetic-based input systems. We propose to improve the robustness of NMT to homophone noise by 1) jointly embedding both textual and phonetic information of source sentences, and 2) augmenting the training dataset with homophone noise. Interestingly, we found that in order to achieve the best translation quality, most (though not all) weights should be put on the phonetic rather than textual information, where the latter is only used as auxiliary information. Experiments show that our method not only significantly improves the robustness of NMT to homophone noise, which is expected but also surprisingly improves the translation quality on clean test sets.
https://arxiv.org/abs/1810.06729
Image captioning is an interdisciplinary research problem that stands between computer vision and natural language processing. The task is to generate a textual description of the content of an image. The typical model used for image captioning is an encoder-decoder deep network, where the encoder captures the essence of an image while the decoder is responsible for generating a sentence describing the image. Attention mechanisms can be used to automatically focus the decoder on parts of the image which are relevant to predict the next word. In this paper, we explore different decoders and attentional models popular in neural machine translation, namely attentional recurrent neural networks, self-attentional transformers, and fully-convolutional networks, which represent the current state of the art of neural machine translation. The image captioning module is available as part of SOCKEYE at this https URL which tutorial can be found at this https URL .
https://arxiv.org/abs/1810.04101
Neural Image Captioning (NIC) or neural caption generation has attracted a lot of attention over the last few years. Describing an image with a natural language has been an emerging challenge in both fields of computer vision and language processing. Therefore a lot of research has focused on driving this task forward with new creative ideas. So far, the goal has been to maximize scores on automated metric and to do so, one has to come up with a plurality of new modules and techniques. Once these add up, the models become complex and resource-hungry. In this paper, we take a small step backwards in order to study an architecture with interesting trade-off between performance and computational complexity. To do so, we tackle every component of a neural captioning model and propose one or more solution that lightens the model overall. Our ideas are inspired by two related tasks: Multimodal and Monomodal Neural Machine Translation.
https://arxiv.org/abs/1810.06245
This paper describes the UMONS solution for the Multimodal Machine Translation Task presented at the third conference on machine translation (WMT18). We explore a novel architecture, called deepGRU, based on recent findings in the related task of Neural Image Captioning (NIC). The models presented in the following sections lead to the best METEOR translation score for both constrained (English, image) -> German and (English, image) -> French sub-tasks.
https://arxiv.org/abs/1810.06233
This report demonstrates our solution for the Open Images 2018 Challenge. Based on our detailed analysis on the Open Images Datasets (OID), it is found that there are four typical features: large-scale, hierarchical tag system, severe annotation incompleteness and data imbalance. Considering these characteristics, an amount of strategies are employed, including SNIPER, soft sampling, class-aware sampling (CAS), hierarchical non-maximum suppression (HNMS) and so on. In virtue of these effective strategies, and further using the powerful SENet154 armed with feature pyramid module and deformable ROIalign as the backbone, our best single model could achieve a mAP of 56.9%. After a further ensemble with 9 models, the final mAP is boosted to 62.2% in the public leaderboard (ranked the 2nd place) and 58.6% in the private leaderboard (ranked the 3rd place, slightly inferior to the 1st place by only 0.04 point).
https://arxiv.org/abs/1810.06208
Attention mechanism, including global attention and local attention, plays a key role in neural machine translation (NMT). Global attention attends to all source words for word prediction. In comparison, local attention selectively looks at fixed-window source words. However, alignment weights for the current target word often decrease to the left and right by linear distance centering on the aligned source position and neglect syntax-directed distance constraints. In this paper, we extend local attention with syntax-distance constraint, to focus on syntactically related source words with the predicted target word, thus learning a more effective context vector for word prediction. Moreover, we further propose a double context NMT architecture, which consists of a global context vector and a syntax-directed context vector over the global attention, to provide more translation performance for NMT from source representation. The experiments on the large-scale Chinese-to-English and English-to-Germen translation tasks show that the proposed approach achieves a substantial and significant improvement over the baseline system.
https://arxiv.org/abs/1711.04231
In Neural Machine Translation (NMT), the decoder can capture the features of the entire prediction history with neural connections and representations. This means that partial hypotheses with different prefixes will be regarded differently no matter how similar they are. However, this might be inefficient since some partial hypotheses can contain only local differences that will not influence future predictions. In this work, we introduce recombination in NMT decoding based on the concept of the “equivalence” of partial hypotheses. Heuristically, we use a simple $n$-gram suffix based equivalence function and adapt it into beam search decoding. Through experiments on large-scale Chinese-to-English and English-to-Germen translation tasks, we show that the proposed method can obtain similar translation quality with a smaller beam size, making NMT decoding more efficient.
https://arxiv.org/abs/1808.08482
Generating a description of an image is called image captioning. Image captioning requires to recognize the important objects, their attributes and their relationships in an image. It also needs to generate syntactically and semantically correct sentences. Deep learning-based techniques are capable of handling the complexities and challenges of image captioning. In this survey paper, we aim to present a comprehensive review of existing deep learning-based image captioning techniques. We discuss the foundation of the techniques to analyze their performances, strengths and limitations. We also discuss the datasets and the evaluation metrics popularly used in deep learning based automatic image captioning.
https://arxiv.org/abs/1810.04020
Weakly Supervised Object Detection (WSOD), using only image-level annotations to train object detectors, is of growing importance in object recognition. In this paper, we propose a novel deep network for WSOD. Unlike previous networks that transfer the object detection problem to an image classification problem using Multiple Instance Learning (MIL), our strategy generates proposal clusters to learn refined instance classifiers by an iterative process. The proposals in the same cluster are spatially adjacent and associated with the same object. This prevents the network from concentrating too much on parts of objects instead of whole objects. We first show that instances can be assigned object or background labels directly based on proposal clusters for instance classifier refinement, and then show that treating each cluster as a small new bag yields fewer ambiguities than the directly assigning label method. The iterative instance classifier refinement is implemented online using multiple streams in convolutional neural networks, where the first is an MIL network and the others are for instance classifier refinement supervised by the preceding one. Experiments are conducted on the PASCAL VOC, ImageNet detection, and MS-COCO benchmarks for WSOD. Results show that our method outperforms the previous state of the art significantly.
https://arxiv.org/abs/1807.03342
Object detection in streaming images is a major step in different detection-based applications, such as object tracking, action recognition, robot navigation, and visual surveillance applications. In mostcases, image quality is noisy and biased, and as a result, the data distributions are disturbed and imbalanced. Most object detection approaches, such as the faster region-based convolutional neural network (Faster RCNN), Single Shot Multibox Detector with 300x300 inputs (SSD300), and You Only Look Once version 2 (YOLOv2), rely on simple sampling without considering distortions and noise under real-world changing environments, despite poor object labeling. In this paper, we propose an Incremental active semi-supervised learning (IASSL) technology for unseen object detection. It combines batch-based active learning (AL) and bin-based semi-supervised learning (SSL) to leverage the strong points of AL’s exploration and SSL’s exploitation capabilities. A collaborative sampling method is also adopted to measure the uncertainty and diversity of AL and the confidence in SSL. Batch-based AL allows us to select more informative, confident, and representative samples with low cost. Bin-based SSL divides streaming image samples into several bins, and each bin repeatedly transfers the discriminative knowledge of convolutional neural network (CNN) deep learning to the next bin until the performance criterion is reached. IASSL can overcome noisy and biased labels in unknown, cluttered data distributions. We obtain superior performance, compared to state-of-the-art technologies such as Faster RCNN, SSD300, and YOLOv2.
https://arxiv.org/abs/1810.10323
Fully convolutional neural networks (FCNs) have shown outstanding performance in many computer vision tasks including salient object detection. However, there still remains two issues needed to be addressed in deep learning based saliency detection. One is the lack of tremendous amount of annotated data to train a network. The other is the lack of robustness for extracting salient objects in images containing complex scenes. In this paper, we present a new architecture$ - $PDNet, a robust prior-model guided depth-enhanced network for RGB-D salient object detection. In contrast to existing works, in which RGB-D values of image pixels are fed directly to a network, the proposed architecture is composed of a master network for processing RGB values, and a sub-network making full use of depth cues and incorporate depth-based features into the master network. To overcome the limited size of the labeled RGB-D dataset for training, we employ a large conventional RGB dataset to pre-train the master network, which proves to contribute largely to the final accuracy. Extensive evaluations over five benchmark datasets demonstrate that our proposed method performs favorably against the state-of-the-art approaches.
https://arxiv.org/abs/1803.08636
In this paper, a novel circular and structural operator tracker (CSOT) is proposed for high performance visual tracking, it not only possesses the powerful discriminative capability of SOSVM but also efficiently inherits the superior computational efficiency of DCF. Based on the proposed circular and structural operators, a set of primal confidence score maps can be obtained by circular correlating feature maps with their corresponding structural correlation filters. Furthermore, an implicit interpolation is applied to convert the multi-resolution feature maps to the continuous domain and make all primal confidence score maps have the same spatial resolution. Then, we exploit an efficient ensemble post-processor based on relative entropy, which can coalesce primal confidence score maps and create an optimal confidence score map for more accurate localization. The target is localized on the peak of the optimal confidence score map. Besides, we introduce a collaborative optimization strategy to update circular and structural operators by iteratively training structural correlation filters, which significantly reduces computational complexity and improves robustness. Experimental results demonstrate that our approach achieves state-of-the-art performance in mean AUC scores of 71.5% and 69.4% on the OTB-2013 and OTB-2015 benchmarks respectively, and obtains a third-best expected average overlap (EAO) score of 29.8% on the VOT-2017 benchmark.
http://arxiv.org/abs/1804.08208
Correlation filter (CF) based tracking algorithms have demonstrated favorable performance recently. Nevertheless, the top performance trackers always employ complicated optimization methods which constraint their real-time applications. How to accelerate the tracking speed while retaining the tracking accuracy is a significant issue. In this paper, we propose a multi-level CF-based tracking approach named MLCFT which further explores the potential capacity of CF with two-stage detection: primal detection and oriented re-detection. The cascaded detection scheme is simple but competent to prevent model drift and accelerate the speed. An effective fusion method based on relative entropy is introduced to combine the complementary features extracted from deep and shallow layers of convolutional neural networks (CNN). Moreover, a novel online model update strategy is utilized in our tracker, which enhances the tracking performance further. Experimental results demonstrate that our proposed approach outperforms the most state-of-the-art trackers while tracking at speed of exceeded 16 frames per second on challenging benchmarks.
http://arxiv.org/abs/1810.05810
Generative Adversarial Networks (GAN) can achieve promising performance on learning complex data distributions on different types of data. In this paper, we first show a straightforward extension of existing GAN algorithm is not applicable to point clouds, because the constraint required for discriminators is undefined for set data. We propose a two fold modification to GAN algorithm for learning to generate point clouds (PC-GAN). First, we combine ideas from hierarchical Bayesian modeling and implicit generative models by learning a hierarchical and interpretable sampling process. A key component of our method is that we train a posterior inference network for the hidden variables. Second, instead of using only state-of-the-art Wasserstein GAN objective, we propose a sandwiching objective, which results in a tighter Wasserstein distance estimate than the commonly used dual form. Thereby, PC-GAN defines a generic framework that can incorporate many existing GAN algorithms. We validate our claims on ModelNet40 benchmark dataset. Using the distance between generated point clouds and true meshes as metric, we find that PC-GAN trained by the sandwiching objective achieves better results on test data than the existing methods. Moreover, as a byproduct, PC- GAN learns versatile latent representations of point clouds, which can achieve competitive performance with other unsupervised learning algorithms on object recognition task. Lastly, we also provide studies on generating unseen classes of objects and transforming image to point cloud, which demonstrates the compelling generalization capability and potentials of PC-GAN.
https://arxiv.org/abs/1810.05795
Domain specific accelerators present new challenges and opportunities for code generation onto novel instruction sets, communication fabrics, and memory architectures. In this paper we introduce an intermediate representation (IR) which enables both deep learning computational kernels and hardware capabilities to be described in the same IR. We then formulate and apply instruction mapping to determine the possible ways a computation can be performed on a hardware system. Next, our scheduler chooses a specific mapping and determines the data movement and computation order. In order to manage the large search space of mappings and schedules, we developed a flexible framework that allows heuristics, cost models, and potentially machine learning to facilitate this search problem. With this system, we demonstrate the automated extraction of matrix multiplication kernels out of recent deep learning kernels such as depthwise-separable convolution. In addition, we demonstrate two to five times better performance on DeepBench sized GEMMs and GRU RNN execution when compared to state-of-the-art (SOTA) implementations on new hardware and up to 85% of the performance for SOTA implementations on existing hardware.
https://arxiv.org/abs/1810.09958
This paper addresses the sensitivity of neural image caption generators to their visual input. A sensitivity analysis and omission analysis based on image foils is reported, showing that the extent to which image captioning architectures retain and are sensitive to visual information varies depending on the type of word being generated and the position in the caption as a whole. We motivate this work in the context of broader goals in the field to achieve more explainability in AI.
https://arxiv.org/abs/1810.05475
Image caption generation systems are typically evaluated against reference outputs. We show that it is possible to predict output quality without generating the captions, based on the probability assigned by the neural model to the reference captions. Such pre-gen metrics are strongly correlated to standard evaluation metrics.
https://arxiv.org/abs/1810.05474
This paper presents the NICT’s participation in the WMT18 shared parallel corpus filtering task. The organizers provided 1 billion words German-English corpus crawled from the web as part of the Paracrawl project. This corpus is too noisy to build an acceptable neural machine translation (NMT) system. Using the clean data of the WMT18 shared news translation task, we designed several features and trained a classifier to score each sentence pairs in the noisy data. Finally, we sampled 100 million and 10 million words and built corresponding NMT systems. Empirical results show that our NMT systems trained on sampled data achieve promising performance.
https://arxiv.org/abs/1809.07043
This paper presents the NICT’s participation to the WMT18 shared news translation task. We participated in the eight translation directions of four language pairs: Estonian-English, Finnish-English, Turkish-English and Chinese-English. For each translation direction, we prepared state-of-the-art statistical (SMT) and neural (NMT) machine translation systems. Our NMT systems were trained with the transformer architecture using the provided parallel data enlarged with a large quantity of back-translated monolingual data that we generated with a new incremental training framework. Our primary submissions to the task are the result of a simple combination of our SMT and NMT systems. Our systems are ranked first for the Estonian-English and Finnish-English language pairs (constraint) according to BLEU-cased.
https://arxiv.org/abs/1809.07037
We propose a novel autoencoding model called Pairwise Augmented GANs. We train a generator and an encoder jointly and in an adversarial manner. The generator network learns to sample realistic objects. In turn, the encoder network at the same time is trained to map the true data distribution to the prior in latent space. To ensure good reconstructions, we introduce an augmented adversarial reconstruction loss. Here we train a discriminator to distinguish two types of pairs: an object with its augmentation and the one with its reconstruction. We show that such adversarial loss compares objects based on the content rather than on the exact match. We experimentally demonstrate that our model generates samples and reconstructions of quality competitive with state-of-the-art on datasets MNIST, CIFAR10, CelebA and achieves good quantitative results on CIFAR10.
https://arxiv.org/abs/1810.04920
This paper studies finding the K nearest neighbors (KNN) of all points in a dataset. Typical solutions to KNN searches use indexing to prune the search, which reduces the number of candidate points that may be within the set of the nearest $K$ points of each query point. In high dimensionality, index searches degrade, making the KNN self-join a prohibitively expensive operation in some scenarios. Furthermore, there are a significant number of distance calculations needed to determine which points are nearest to each query point. To address these challenges, we propose a hybrid CPU/GPU approach. Since the CPU and GPU are considerably different architectures that are best exploited using different algorithms, we advocate for splitting the work between both architectures based on the characteristic workloads defined by the query points in the dataset. As such, we assign dense regions to the GPU, and sparse regions to the CPU to most efficiently exploit the relative strengths of each architecture. Critically, we find that the relative performance gains over the reference implementation across four real-world datasets are a function of the data properties (size, dimensionality, distribution), and number of neighbors, K.
https://arxiv.org/abs/1810.04758
Building on the success of deep learning, two modern approaches to learn a probability model of the observed data are Generative Adversarial Networks (GANs) and Variational AutoEncoders (VAEs). VAEs consider an explicit probability model for the data and compute a generative distribution by maximizing a variational lower-bound on the log-likelihood function. GANs, however, compute a generative model by minimizing a distance between observed and generated probability distributions without considering an explicit model for the observed data. The lack of having explicit probability models in GANs prohibits computation of sample likelihoods in their frameworks and limits their use in statistical inference problems. In this work, we show that an optimal transport GAN with the entropy regularization can be viewed as a generative model that maximizes a lower-bound on average sample likelihoods, an approach that VAEs are based on. In particular, our proof constructs an explicit probability model for GANs that can be used to compute likelihood statistics within GAN’s framework. Our numerical results on several datasets demonstrate consistent trends with the proposed theory.
https://arxiv.org/abs/1810.04147
We propose UOLO, a novel framework for the simultaneous detection and segmentation of structures of interest in medical images. UOLO consists of an object segmentation module which intermediate abstract representations are processed and used as input for object detection. The resulting system is optimized simultaneously for detecting a class of objects and segmenting an optionally different class of structures. UOLO is trained on a set of bounding boxes enclosing the objects to detect, as well as pixel-wise segmentation information, when available. A new loss function is devised, taking into account whether a reference segmentation is accessible for each training image, in order to suitably backpropagate the error. We validate UOLO on the task of simultaneous optic disc (OD) detection, fovea detection, and OD segmentation from retinal images, achieving state-of-the-art performance on public datasets.
https://arxiv.org/abs/1810.05729
This work investigates an alternative model for neural machine translation (NMT) and proposes a novel architecture, where we employ a multi-dimensional long short-term memory (MDLSTM) for translation modeling. In the state-of-the-art methods, source and target sentences are treated as one-dimensional sequences over time, while we view translation as a two-dimensional (2D) mapping using an MDLSTM layer to define the correspondence between source and target words. We extend beyond the current sequence to sequence backbone NMT models to a 2D structure in which the source and target sentences are aligned with each other in a 2D grid. Our proposed topology shows consistent improvements over attention-based sequence to sequence model on two WMT 2017 tasks, German$\leftrightarrow$English.
https://arxiv.org/abs/1810.03975
Many biological, social and man-made systems are better described in terms of temporal networks, i.e. networks whose links are only present at certain points in time, rather than by static ones. In particular, it has been found that non-Markovianity is a necessary ingredient to capture the non-trivial temporal patterns of real-world networks. However, our understanding of how memory can affect the properties of dynamical processes taking place over temporal networks is still very limited, being especially constrained to the case of short-term memory. Here, by introducing a model for temporal networks in which we can precisely control the link density and the strength and length of memory for each link, we unveil the role played by memory on the dynamics of epidemic spreading processes. Surprisingly, we find that the average spreading time in our temporal networks is often non-monotonically dependent on the length of the memory, and that the optimal value of the memory length which maximizes the spreading time depends on the strength of the memory and on the density of links in the network. Through analytical arguments we then explore the effect that changing the number and length of network paths connecting any two nodes has on the value of optimal memory.
https://arxiv.org/abs/1810.05240
Attention mechanisms have been widely used in Visual Question Answering (VQA) solutions due to their capacity to model deep cross-domain interactions. Analyzing attention maps offers us a perspective to find out limitations of current VQA systems and an opportunity to further improve them. In this paper, we select two state-of-the-art VQA approaches with attention mechanisms to study their robustness and disadvantages by visualizing and analyzing their estimated attention maps. We find that both methods are sensitive to features, and simultaneously, they perform badly for counting and multi-object related questions. We believe that the findings and analytical method will help researchers identify crucial challenges on the way to improve their own VQA systems.
https://arxiv.org/abs/1810.03821
We present detailed Raman studies of graphene deposited on gallium nitride nanowires with different variations in height. Our results show that different density and height of nanowires being in contact with graphene impact graphene properties like roughness, strain and carrier concentration as well as density and type of induced defects. Detailed analysis of Raman spectra of graphene deposited on different nanowire substrates shows that bigger differences in nanowires height increase graphene strain, while higher number of nanowires in contact with graphene locally reduce the strain. Moreover, the value of graphene carrier concentration is found to be correlated with the density of nanowires in contact with graphene. Analysis of intensity ratios of Raman G, D and D’ bands enable to trace how nanowire substrate impacts the defect concentration and type. The lowest concentration of defects is observed for graphene deposited on nanowires of the lowest density. Contact between graphene and densely arranged nanowires leads to a large density of vacancies. On the other hand, grain boundaries are the main type of defects in graphene on rarely distributed nanowires. Our results also show modification of graphene carrier concentration and strain by different types of defects present in graphene.
https://arxiv.org/abs/1810.03668
In microblog retrieval, query expansion can be essential to obtain good search results due to the short size of queries and posts. Since information in microblogs is highly dynamic, an up-to-date index coupled with pseudo-relevance feedback (PRF) with an external corpus has a higher chance of retrieving more relevant documents and improving ranking. In this paper, we focus on the research question:how can we reduce the query expansion computational cost while maintaining the same retrieval precision as standard PRF? Therefore, we propose to accelerate the query expansion step of pseudo-relevance feedback. The hypothesis is that using an expansion corpus organized into verticals for expanding the query, will lead to a more efficient query expansion process and improved retrieval effectiveness. Thus, the proposed query expansion method uses a distributed search architecture and resource selection algorithms to provide an efficient query expansion process. Experiments on the TREC Microblog datasets show that the proposed approach can match or outperform standard PRF in MAP and NDCG@30, with a computational cost that is three orders of magnitude lower.
https://arxiv.org/abs/1810.03519
Motion artifacts are a primary source of magnetic resonance (MR) image quality deterioration with strong repercussions on diagnostic performance. Currently, MR motion correction is carried out either prospectively, with the help of motion tracking systems, or retrospectively by mainly utilizing computationally expensive iterative algorithms. In this paper, we utilize a new adversarial framework, titled MedGAN, for the joint retrospective correction of rigid and non-rigid motion artifacts in different body regions and without the need for a reference image. MedGAN utilizes a unique combination of non-adversarial losses and a new generator architecture to capture the textures and fine-detailed structures of the desired artifact-free MR images. Quantitative and qualitative comparisons with other adversarial techniques have illustrated the proposed model performance.
https://arxiv.org/abs/1809.06276
Text mining and analytics software has become popular, but little attention has been paid to the software architectures of such systems. Often they are built from scratch using special-purpose software and data structures, which increases their cost and complexity. This demo paper describes Sifaka, a new open-source text mining application constructed above a standard search engine index using existing application programmer interface (API) calls. Indexing integrates popular annotation software libraries to augment the full-text index with noun phrase and named-entities; n-grams are also provided. Sifaka enables a person to quickly explore and analyze large text collections using search, frequency analysis, and co-occurrence analysis; and import existing document labels or interactively construct document sets that are positive or negative examples of new concepts, perform feature selection, and export feature vectors compatible with popular machine learning software. Sifaka demonstrates that search engines are good platforms for text mining applications while also making common IR text mining capabilities accessible to researchers in disciplines where programming skills are less common.
https://arxiv.org/abs/1810.02907
In this paper, we analyze failure cases of state-of-the-art detectors and observe that most hard false positives result from classification instead of localization and they have a large negative impact on the performance of object detectors. We conjecture there are three factors: (1) Shared feature representation is not optimal due to the mismatched goals of feature learning for classification and localization; (2) multi-task learning helps, yet optimization of the multi-task loss may result in sub-optimal for individual tasks; (3) large receptive field for different scales leads to redundant context information for small objects. We demonstrate the potential of detector classification power by a simple, effective, and widely-applicable Decoupled Classification Refinement (DCR) network. In particular, DCR places a separate classification network in parallel with the localization network (base detector). With ROI Pooling placed on the early stage of the classification network, we enforce an adaptive receptive field in DCR. During training, DCR samples hard false positives from the base detector and trains a strong classifier to refine classification results. During testing, DCR refines all boxes from the base detector. Experiments show competitive results on PASCAL VOC and COCO without any bells and whistles. Our codes are available at: this https URL.
https://arxiv.org/abs/1810.04002
We propose a new recurrent generative model for generating images from text captions while attending on specific parts of text captions. Our model creates images by incrementally adding patches on a “canvas” while attending on words from text caption at each timestep. Finally, the canvas is passed through an upscaling network to generate images. We also introduce a new method for generating visual-semantic sentence embeddings based on self-attention over text. We compare our model’s generated images with those generated Reed et. al.’s model and show that our model is a stronger baseline for text to image generation tasks.
https://arxiv.org/abs/1810.02833
This paper demonstrates that word sense disambiguation (WSD) can improve neural machine translation (NMT) by widening the source context considered when modeling the senses of potentially ambiguous words. We first introduce three adaptive clustering algorithms for WSD, based on k-means, Chinese restaurant processes, and random walks, which are then applied to large word contexts represented in a low-rank space and evaluated on SemEval shared-task data. We then learn word vectors jointly with sense vectors defined by our best WSD method, within a state-of-the-art NMT system. We show that the concatenation of these vectors, and the use of a sense selection mechanism based on the weighted average of sense vectors, outperforms several baselines including sense-aware ones. This is demonstrated by translation on five language pairs. The improvements are above one BLEU point over strong NMT baselines, +4% accuracy over all ambiguous nouns and verbs, or +20% when scored manually over several challenging words.
https://arxiv.org/abs/1810.02614
We propose a method for the weakly supervised detection of objects in paintings. At training time, only image-level annotations are needed. This, combined with the efficiency of our multiple-instance learning method, enables one to learn new classes on-the-fly from globally annotated databases, avoiding the tedious task of manually marking objects. We show on several databases that dropping the instance-level annotations only yields mild performance losses. We also introduce a new database, IconArt, on which we perform detection experiments on classes that could not be learned on photographs, such as Jesus Child or Saint Sebastian. To the best of our knowledge, these are the first experiments dealing with the automatic (and in our case weakly supervised) detection of iconographic elements in paintings. We believe that such a method is of great benefit for helping art historians to explore large digital databases.
https://arxiv.org/abs/1810.02569
Wasserstein GAN(WGAN) is a model that minimizes the Wasserstein distance between a data distribution and sample distribution. Recent studies have proposed stabilizing the training process for the WGAN and implementing the Lipschitz constraint. In this study, we prove the local stability of optimizing the simple gradient penalty $\mu$-WGAN(SGP $\mu$-WGAN) under suitable assumptions regarding the equilibrium and penalty measure $\mu$. The measure valued differentiation concept is employed to deal with the derivative of the penalty terms, which is helpful for handling abstract singular measures with lower dimensional support. Based on this analysis, we claim that penalizing the data manifold or sample manifold is the key to regularizing the original WGAN with a gradient penalty. Experimental results obtained with unintuitive penalty measures that satisfy our assumptions are also provided to support our theoretical results.
https://arxiv.org/abs/1810.02528
Many machine learning problems involve iteratively and alternately optimizing different task objectives with respect to different sets of parameters. Appropriately scheduling the optimization of a task objective or a set of parameters is usually crucial to the quality of convergence. In this paper, we present AutoLoss, a meta-learning framework that automatically learns and determines the optimization schedule. AutoLoss provides a generic way to represent and learn the discrete optimization schedule from metadata, allows for a dynamic and data-driven schedule in ML problems that involve alternating updates of different parameters or from different loss objectives. We apply AutoLoss on four ML tasks: d-ary quadratic regression, classification using a multi-layer perceptron (MLP), image generation using GANs, and multi-task neural machine translation (NMT). We show that the AutoLoss controller is able to capture the distribution of better optimization schedules that result in higher quality of convergence on all four tasks. The trained AutoLoss controller is generalizable – it can guide and improve the learning of a new task model with different specifications, or on different datasets.
https://arxiv.org/abs/1810.02442
Within ten years, the era of large-scale systematics surveys will decay thanks to a complete census of exoplanetary systems within 200 pc from the Sun. With the first Lights foreseen between 2024 and 2028, the new generation of extremely large telescopes and planet imagers will arrive at a propitious time to exploit this manna of discoveries to characterize the formation, the evolution, and the physics of giant and telluric planets with the ultimate goal to search and discover bio-signatures. In that perspective, I will briefly summarize the main characteristics of the direct imaging instruments of the ELTs dedicated to the study of exoplanets, and I will review the key science cases (from the initial conditions of planetary formation, the architecture of planetary systems and the physics and atmospheres of giant and telluric planets) that they will address given their predicted performances.
https://arxiv.org/abs/1810.02031
There is growing interest in object detection in advanced driver assistance systems and autonomous robots and vehicles. To enable such innovative systems, we need faster object detection. In this work, we investigate the trade-off between accuracy and speed with domain-specific approximations, i.e. category-aware image size scaling and proposals scaling, for two state-of-the-art deep learning-based object detection meta-architectures. We study the effectiveness of applying approximation both statically and dynamically to understand the potential and the applicability of them. By conducting experiments on the ImageNet VID dataset, we show that domain-specific approximation has great potential to improve the speed of the system without deteriorating the accuracy of object detectors, i.e. up to 7.5x speedup for dynamic domain-specific approximation. To this end, we present our insights toward harvesting domain-specific approximation as well as devise a proof-of-concept runtime, AutoFocus, that exploits dynamic domain-specific approximation.
https://arxiv.org/abs/1810.02010
Neural Machine Translation (NMT) has been widely adopted recently due to its advantages compared with the traditional Statistical Machine Translation (SMT). However, an NMT system still often produces translation failures due to the complexity of natural language and sophistication in designing neural networks. While in-house black-box system testing based on reference translations (i.e., examples of valid translations) has been a common practice for NMT quality assurance, an increasingly critical industrial practice, named in-vivo testing, exposes unseen types or instances of translation failures when real users are using a deployed industrial NMT system. To fill the gap of lacking test oracle for in-vivo testing of an NMT system, in this paper, we propose a new approach for automatically identifying translation failures, without requiring reference translations for a translation task; our approach can directly serve as a test oracle for in-vivo testing. Our approach focuses on properties of natural language translation that can be checked systematically and uses information from both the test inputs (i.e., the texts to be translated) and the test outputs (i.e., the translations under inspection) of the NMT system. Our evaluation conducted on real-world datasets shows that our approach can effectively detect targeted property violations as translation failures. Our experiences on deploying our approach in both production and development environments of WeChat (a messenger app with over one billion monthly active users) demonstrate high effectiveness of our approach along with high industry impact.
https://arxiv.org/abs/1807.02340