This short paper presents the design decisions taken and challenges encountered in completing SemEval Task 6, which poses the problem of identifying and categorizing offensive language in tweets. Our proposed solutions explore Deep Learning techniques, Linear Support Vector classification and Random Forests to identify offensive tweets, to classify offenses as targeted or untargeted and eventually to identify the target subject type.
http://arxiv.org/abs/1903.00665
Estimating unknown rotations from noisy measurements is an important step in SfM and other 3D vision tasks. Typically, local optimization methods susceptible to returning suboptimal local minima are used to solve the rotation averaging problem. A new wave of approaches that leverage convex relaxations have provided the first formal guarantees of global optimality for state estimation techniques involving SO(3). However, most of these guarantees are only applicable when the measurement error introduced by noise is within a certain bound that depends on the problem instance’s structure. In this paper, we cast rotation averaging as a polynomial optimization problem over unit quaternions to produce the first rotation averaging method that is formally guaranteed to provide a certifiably globally optimal solution for \textit{any} problem instance. This is achieved by formulating and solving a sparse convex sum of squares (SOS) relaxation of the problem. We provide an open source implementation of our algorithm and experiments, demonstrating the benefits of our globally optimal approach.
https://arxiv.org/abs/1904.01645
Plant Phenomics based on imaging based techniques can be used to monitor the health and the diseases of plants and crops. The use of 3D data for plant phenomics is a recent phenomenon. However, since 3D point cloud contains more information than plant images, in this paper, we compare the performance of different keypoint detectors and local feature descriptors combinations for the plant growth stage and it’s growth condition classification based on 3D point clouds of the plants. We have also implemented a modified form of 3D SIFT descriptor, that is invariant to rotation and is computationally less intense than most of the 3D SIFT descriptors reported in the existing literature. The performance is evaluated in terms of the classification accuracy and the results are presented in terms of accuracy tables. We find the ISS-SHOT and the SIFT-SIFT combinations consistently perform better and Fisher Vector (FV) is a better encoder than Vector of Linearly Aggregated (VLAD) for such applications. It can serve as a better modality.
http://arxiv.org/abs/1904.08493
This work provides a strong baseline for the problem of multi-source multi-target domain adaptation and generalization in medical imaging. Using a diverse collection of ten chest X-ray datasets, we empirically demonstrate the benefits of training medical imaging deep learning models on varied patient populations for generalization to out-of-sample domains.
https://arxiv.org/abs/1904.01638
In many cases, especially with medical images, it is prohibitively challenging to produce a sufficiently large training sample of pixel-level annotations to train deep neural networks for semantic image segmentation. On the other hand, some information is often known about the contents of images. We leverage information on whether an image presents the segmentation target or whether it is absent from the image to improve segmentation performance by augmenting the amount of data usable for model training. Specifically, we propose a semi-supervised framework that employs image-to-image translation between weak labels (e.g., presence vs. absence of cancer), in addition to fully supervised segmentation on some examples. We conjecture that this translation objective is well aligned with the segmentation objective as both require the same disentangling of image variations. Building on prior image-to-image translation work, we re-use the encoder and decoders for translating in either direction between two domains, employing a strategy of selectively decoding domain-specific variations. For presence vs. absence domains, the encoder produces variations that are common to both and those unique to the presence domain. Furthermore, we successfully re-use one of the decoders used in translation for segmentation. We validate the proposed method on synthetic tasks of varying difficulty as well as on the real task of brain tumor segmentation in magnetic resonance images, where we show significant improvements over standard semi-supervised training with autoencoding.
https://arxiv.org/abs/1904.01636
Social scientists have recently turned to analyzing text using tools from natural language processing like word embeddings to measure concepts like ideology, bias, and affinity. However, word embeddings are difficult to use in the regression framework familiar to social scientists: embeddings are are neither identified, nor directly interpretable. I offer two advances on standard embedding models to remedy these problems. First, I develop Bayesian Word Embeddings with Automatic Relevance Determination priors, relaxing the assumption that all embedding dimensions have equal weight. Second, I apply work identifying latent variable models to anchor the dimensions of the resulting embeddings, identifying them, and making them interpretable and usable in a regression. I then apply this model and anchoring approach to two cases, the shift in internationalist rhetoric in the American presidents’ inaugural addresses, and the relationship between bellicosity in American foreign policy decision-makers’ deliberations. I find that inaugural addresses became less internationalist after 1945, which goes against the conventional wisdom, and that an increase in bellicosity is associated with an increase in hostile actions by the United States, showing that elite deliberations are not cheap talk, and helping confirm the validity of the model.
https://arxiv.org/abs/1904.01628
This is a report of our lessons learned building acoustic models from 1 Million hours of unlabeled speech, while labeled speech is restricted to 7,000 hours. We employ student/teacher training on unlabeled data, helping scale out target generation in comparison to confidence model based methods, which require a decoder and a confidence model. To optimize storage and to parallelize target generation, we store high valued logits from the teacher model. Introducing the notion of scheduled learning, we interleave learning on unlabeled and labeled data. To scale distributed training across a large number of GPUs, we use BMUF with 64 GPUs, while performing sequence training only on labeled data with gradient threshold compression SGD using 16 GPUs. Our experiments show that extremely large amounts of data are indeed useful; with little hyper-parameter tuning, we obtain relative WER improvements in the 10 to 20% range, with higher gains in noisier conditions.
https://arxiv.org/abs/1904.01624
Gait based biometric aims to discriminate among people by the way or manner they walk. It represents a biometric at distance which has many advantages over other biometric modalities. State-of-the-art methods require a limited cooperation from the individuals. Consequently, contrary to other modalities, gait is a non-invasive approach. As a behavioral analysis, gait is difficult to circumvent. Moreover, gait can be performed without the subject being aware of it. Consequently, it is more difficult to try to tamper one own biometric signature. In this paper we review different features and approaches used in gait recognition. A novel method able to learn the discriminative human body-parts to improve the recognition accuracy will be introduced. Extensive experiments will be performed on CASIA gait benchmark database and results will be compared to state-of-the-art methods.
https://arxiv.org/abs/1904.01620
Learning high-quality embeddings for rare words is a hard problem because of sparse context information. Mimicking (Pinter et al., 2017) has been proposed as a solution: given embeddings learned by a standard algorithm, a model is first trained to reproduce embeddings of frequent words from their surface form and then used to compute embeddings for rare words. In this paper, we introduce attentive mimicking: the mimicking model is given access not only to a word’s surface form, but also to all available contexts and learns to attend to the most informative and reliable contexts for computing an embedding. In an evaluation on four tasks, we show that attentive mimicking outperforms previous work for both rare and medium-frequency words. Thus, compared to previous work, attentive mimicking improves embeddings for a much larger part of the vocabulary, including the medium-frequency range.
https://arxiv.org/abs/1904.01617
Identifying the intent of a citation in scientific papers (e.g., background information, use of methods, comparing results) is critical for machine reading of individual publications and automated analysis of the scientific literature. We propose structural scaffolds, a multitask model to incorporate structural information of scientific papers into citations for effective classification of citation intents. Our model achieves a new state-of-the-art on an existing ACL anthology dataset (ACL-ARC) with a 13.3% absolute increase in F1 score, without relying on external linguistic resources or hand-engineered features as done in existing methods. In addition, we introduce a new dataset of citation intents (SciCite) which is more than five times larger and covers multiple scientific domains compared with existing datasets. Our code and data are available at: this https URL.
https://arxiv.org/abs/1904.01608
Electroencephalography (EEG) recordings of brain activity taken while participants read or listen to language are widely used within the cognitive neuroscience and psycholinguistics communities as a tool to study language comprehension. Several time-locked stereotyped EEG responses to word-presentations – known collectively as event-related potentials (ERPs) – are thought to be markers for semantic or syntactic processes that take place during comprehension. However, the characterization of each individual ERP in terms of what features of a stream of language trigger the response remains controversial. Improving this characterization would make ERPs a more useful tool for studying language comprehension. We take a step towards better understanding the ERPs by fine-tuning a language model to predict them. This new approach to analysis shows for the first time that all of the ERPs are predictable from embeddings of a stream of language. Prior work has only found two of the ERPs to be predictable. In addition to this analysis, we examine which ERPs benefit from sharing parameters during joint training. We find that two pairs of ERPs previously identified in the literature as being related to each other benefit from joint training, while several other pairs of ERPs that benefit from joint training are suggestive of potential relationships. Extensions of this analysis that further examine what kinds of information in the model embeddings relate to each ERP have the potential to elucidate the processes involved in human language comprehension.
http://arxiv.org/abs/1904.01548
How do we know if a particular medical treatment actually works? Ideally one would consult all available evidence from relevant clinical trials. Unfortunately, such results are primarily disseminated in natural language scientific articles, imposing substantial burden on those trying to make sense of them. In this paper, we present a new task and corpus for making this unstructured evidence actionable. The task entails inferring reported findings from a full-text article describing a randomized controlled trial (RCT) with respect to a given intervention, comparator, and outcome of interest, e.g., inferring if an article provides evidence supporting the use of aspirin to reduce risk of stroke, as compared to placebo. We present a new corpus for this task comprising 10,000+ prompts coupled with full-text articles describing RCTs. Results using a suite of models — ranging from heuristic (rule-based) approaches to attentive neural architectures — demonstrate the difficulty of the task, which we believe largely owes to the lengthy, technical input texts. To facilitate further work on this important, challenging problem we make the corpus, documentation, a website and leaderboard, and code for baselines and evaluation available at this http URL.
https://arxiv.org/abs/1904.01606
We provide an NLP framework to uncover four linguistic dimensions of political polarization in social media: topic choice, framing, affect and illocutionary force. We quantify these aspects with existing lexical methods, and propose clustering of tweet embeddings as a means to identify salient topics for analysis across events; human evaluations show that our approach generates more cohesive topics than traditional LDA-based models. We apply our methods to study 4.4M tweets on 21 mass shootings. We provide evidence that the discussion of these events is highly polarized politically and that this polarization is primarily driven by partisan differences in framing rather than topic choice. We identify framing devices, such as grounding and the contrasting use of the terms “terrorist” and “crazy”, that contribute to polarization. Results pertaining to topic choice, affect and illocutionary force suggest that Republicans focus more on the shooter and event-specific facts (news) while Democrats focus more on the victims and call for policy changes. Our work contributes to a deeper understanding of the way group divisions manifest in language and to computational methods for studying them.
https://arxiv.org/abs/1904.01596
The origin of the observed diversity of planetary system architectures is one of the main topic of the exoplanetary research. The detection of a statistically significant sample of planets around young stars allows us to study the early stages of planet formation and evolution, but only a handful of them is known so far. In this regard, a considerable contribution is expected from the NASA TESS satellite, which is now performing a survey of $\sim 85 \%$ of the sky to search for short-period transiting planets In its first month of operations, TESS found a planet candidate with an orbital period of 8.14 days around a member of the Tuc-Hor young association ($\sim$ 40 Myr), the G6V main component of the binary system DS\,Tuc. If confirmed, it would be the first transiting planet around a young star suitable for radial velocity and/or atmospheric characterization. We aim to validate the planetary nature of this companion and to measure its orbital and physical parameters. We obtain accurate planet parameters by coupling an independent reprocessing of the TESS light curve with improved stellar parameters and the dilution caused by the binary companion; we analyse high precision archival radial velocities to impose an upper limit of 0.94 M${\rm Jup}$ on the planet mass; we finally rule out the presence of external companions beyond 40 au with adaptive optics images. We confirm the presence of a young, giant (${\rm R} = 0.50$ R${\rm Jup}$) and possibly inflated planet (having a theoretical mass lower than 20 M$_{\oplus}$) around DS\,Tuc A. We discuss the feasibility of mass determination, Rossiter-McLaughlin analysis and atmosphere characterization, allowed by the brightness of the star.
https://arxiv.org/abs/1904.01591
People often share personal narratives in order to seek advice from others. To properly infer the narrator’s intention, one needs to apply a certain degree of common sense and social intuition. To test the capabilities of NLP systems to recover such intuition, we introduce the new task of inferring what is the advice-seeking goal behind a personal narrative. We formulate this as a cloze test, where the goal is to identify which of two advice-seeking questions was removed from a given narrative. The main challenge in constructing this task is finding pairs of semantically plausible advice-seeking questions for given narratives. To address this challenge, we devise a method that exploits commonalities in experiences people share online to automatically extract pairs of questions that are appropriate candidates for the cloze task. This results in a dataset of over 20,000 personal narratives, each matched with a pair of related advice-seeking questions: one actually intended by the narrator, and the other one not. The dataset covers a very broad array of human experiences, from dating, to career options, to stolen iPads. We use human annotation to determine the degree to which the task relies on common sense and social intuition in addition to a semantic understanding of the narrative. By introducing several baselines for this new task we demonstrate its feasibility and identify avenues for better modeling the intention of the narrator.
https://arxiv.org/abs/1904.01587
Neural networks for image recognition have evolved through extensive manual design from simple chain-like models to structures with multiple wiring paths. The success of ResNets and DenseNets is due in large part to their innovative wiring plans. Now, neural architecture search (NAS) studies are exploring the joint optimization of wiring and operation types, however, the space of possible wirings is constrained and still driven by manual design despite being searched. In this paper, we explore a more diverse set of connectivity patterns through the lens of randomly wired neural networks. To do this, we first define the concept of a stochastic network generator that encapsulates the entire network generation process. Encapsulation provides a unified view of NAS and randomly wired networks. Then, we use three classical random graph models to generate randomly wired graphs for networks. The results are surprising: several variants of these random generators yield network instances that have competitive accuracy on the ImageNet benchmark. These results suggest that new efforts focusing on designing better network generators may lead to new breakthroughs by exploring less constrained search spaces with more room for novel design.
http://arxiv.org/abs/1904.01569
The Zipf’s law establishes that if the words of a (large) text are ordered by decreasing frequency, the frequency versus the rank decreases as a power law with exponent close to $-1$. Previous work has stressed that this pattern arises from a conflict of interests of the participants of communication. The challenge here is to define a computational multi-agent language game, mainly based on a parameter that measures the relative participant’s interests. Numerical simulations suggest that at critical values of the parameter a human-like vocabulary, exhibiting scaling properties, seems to appear. The appearance of an intermediate distribution of frequencies at some critical values of the parameter suggests that on a population of artificial agents the emergence of scaling partly arises as a self-organized process only from local interactions between agents.
http://arxiv.org/abs/1705.05762
We propose a novel learning paradigm for Deep Neural Networks (DNN) by using Boolean logic algebra. We first present the basic differentiable operators of a Boolean system such as conjunction, disjunction and exclusive-OR and show how these elementary operators can be combined in a simple and meaningful way to form Neural Logic Networks (NLNs). We examine the effectiveness of the proposed NLN framework in learning Boolean functions and discrete-algorithmic tasks. We demonstrate that, in contrast to the implicit learning in MLP approach, the proposed neural logic networks can learn the logical functions explicitly that can be verified and interpreted by human. In particular, we propose a new framework for learning the inductive logic programming (ILP) problems by exploiting the explicit representational power of NLN. We show the proposed neural ILP solver is capable of feats such as predicate invention and recursion and can outperform the current state of the art neural ILP solvers using a variety of benchmark tasks such as decimal addition and multiplication, and sorting on ordered list.
http://arxiv.org/abs/1904.01554
This paper proposes an architecture to enable Web service providers to interact with personal services. Personal services are vanilla HTTP services that are invoked from a browser, upon a request made by a service Provider, to deliver some service on the client side, i.e., on an execution environment defined by the browser’s user. Personal services can be used both to handle content manipulation and presentation or to deliver request-response interactions with different goals (e.g. user authentication). Unlike plugins, that are described to service providers on each and every HTTP request, personal services are explicitly searched by service providers using a novel agent, a Broker, that works in close cooperation with each browser. We have implemented this architecture and implemented an HTTP proxy to cope with it. For demonstration purposes we show how we can use personal services for personal authentication with an electronic identification (eID) card
https://arxiv.org/abs/1904.01541
In the light of ongoing progresses of research on artificial intelligent systems exhibiting a steadily increasing problem-solving ability, the identification of practicable solutions to the value alignment problem in AGI Safety is becoming a matter of urgency. In this context, one preeminent challenge that has been addressed by multiple researchers is the adequate formulation of utility functions or equivalents reliably capturing human ethical conceptions. However, the specification of suitable utility functions harbors the risk of “perverse instantiation” for which no final consensus on responsible proactive countermeasures has been achieved so far. Amidst this background, we propose a novel socio-technological ethical framework denoted Augmented Utilitarianism which directly alleviates the perverse instantiation problem. We elaborate on how augmented by AI and more generally science and technology, it might allow a society to craft and update ethical utility functions while jointly undergoing a dynamical ethical enhancement. Further, we elucidate the need to consider embodied simulations in the design of utility functions for AGIs aligned with human values. Finally, we discuss future prospects regarding the usage of the presented scientifically grounded ethical framework and mention possible challenges.
http://arxiv.org/abs/1904.01540
Removing rain streaks from a single image has been drawing considerable attention as rain streaks can severely degrade the image quality and affect the performance of existing outdoor vision tasks. While recent CNN-based derainers have reported promising performances, deraining remains an open problem for two reasons. First, existing synthesized rain datasets have only limited realism, in terms of modeling real rain characteristics such as rain shape, direction and intensity. Second, there are no public benchmarks for quantitative comparisons on real rain images, which makes the current evaluation less objective. The core challenge is that real world rain/clean image pairs cannot be captured at the same time. In this paper, we address the single image rain removal problem in two ways. First, we propose a semi-automatic method that incorporates temporal priors and human supervision to generate a high-quality clean image from each input sequence of real rain images. Using this method, we construct a large-scale dataset of $\sim$$29.5K$ rain/rain-free image pairs that covers a wide range of natural rain scenes. Second, to better cover the stochastic distribution of real rain streaks, we propose a novel SPatial Attentive Network (SPANet) to remove rain streaks in a local-to-global manner. Extensive experiments demonstrate that our network performs favorably against the state-of-the-art deraining methods.
http://arxiv.org/abs/1904.01538
This work proposes the use of clean speech vocoder parameters as the target for a neural network performing speech enhancement. These parameters have been designed for text-to-speech synthesis so that they both produce high-quality resyntheses and also are straightforward to model with neural networks, but have not been utilized in speech enhancement until now. In comparison to a matched text-to-speech system that is given the ground truth transcripts of the noisy speech, our model is able to produce more natural speech because it has access to the true prosody in the noisy speech. In comparison to two denoising systems, the oracle Wiener mask and a DNN-based mask predictor, our model equals the oracle Wiener mask in subjective quality and intelligibility and surpasses the realistic system. A vocoder-based upper bound shows that there is still room for improvement with this approach beyond the oracle Wiener mask. We test speaker-dependence with two speakers and show that a single model can be used for multiple speakers.
http://arxiv.org/abs/1904.01537
Traditional visual speech recognition systems consist of two stages, feature extraction and classification. Recently, several deep learning approaches have been presented which automatically extract features from the mouth images and aim to replace the feature extraction stage. However, research on joint learning of features and classification remains limited. In addition, most of the existing methods require large amounts of data in order to achieve state-of-the-art performance, otherwise they under-perform. In this work, we present an end-to-end visual speech recognition system based on fully-connected layers and Long-Short Memory (LSTM) networks which is suitable for small-scale datasets. The model consists of two streams which extract features directly from the mouth and difference images, respectively. The temporal dynamics in each stream are modelled by a Bidirectional LSTM (BLSTM) and the fusion of the two streams takes place via another BLSTM. An absolute improvement of 0.6%, 3.4%, 3.9%, 11.4% over the state-of-the-art is reported on the OuluVS2, CUAVE, AVLetters and AVLetters2 databases, respectively.
https://arxiv.org/abs/1904.01954
Facial expression analysis based on machine learning requires large number of well-annotated data to reflect different changes in facial motion. Publicly available datasets truly help to accelerate research in this area by providing a benchmark resource, but all of these datasets, to the best of our knowledge, are limited to rough annotations for action units, including only their absence, presence, or a five-level intensity according to the Facial Action Coding System. To meet the need for videos labeled in great detail, we present a well-annotated dataset named FEAFA for Facial Expression Analysis and 3D Facial Animation. One hundred and twenty-two participants, including children, young adults and elderly people, were recorded in real-world conditions. In addition, 99,356 frames were manually labeled using Expression Quantitative Tool developed by us to quantify 9 symmetrical FACS action units, 10 asymmetrical (unilateral) FACS action units, 2 symmetrical FACS action descriptors and 2 asymmetrical FACS action descriptors, and each action unit or action descriptor is well-annotated with a floating point number between 0 and 1. To provide a baseline for use in future research, a benchmark for the regression of action unit values based on Convolutional Neural Networks are presented. We also demonstrate the potential of our FEAFA dataset for 3D facial animation. Almost all state-of-the-art algorithms for facial animation are achieved based on 3D face reconstruction. We hence propose a novel method that drives virtual characters only based on action unit value regression of the 2D video frames of source actors.
http://arxiv.org/abs/1904.01509
Irony and sarcasm are two complex linguistic phenomena that are widely used in everyday language and especially over the social media, but they represent two serious issues for automated text understanding. Many labeled corpora have been extracted from several sources to accomplish this task, and it seems that sarcasm is found in different ways for different domains. Nonetheless, very little work has been done for comparing different methods among the available corpora. Furthermore, usually, each author extracts and uses its own dataset to train and/or test his own method. In this paper we show that sarcasm detection can be tackled by applying classical machine learning algorithms to input texts sub-symbolically represented in a Latent Semantic space. The main consequence is that our studies establish both reference datasets and baselines for the sarcasm detection problem that could serve to the scientific community in order to test newly proposed method.
http://arxiv.org/abs/1904.04019
Guided super-resolution is a unifying framework for several computer vision tasks where the inputs are a low-resolution source image of some target quantity (e.g., perspective depth acquired with a time-of-flight camera) and a high-resolution guide image from a different domain (e.g., a gray-scale image from a conventional camera); and the target output is a high-resolution version of the source (in our example, a high-res depth map). The standard way of looking at this problem is to formulate it as a super-resolution task, i.e., the source image is upsampled to the target resolution, while transferring the missing high-frequency details from the guide. Here, we propose to turn that interpretation on its head and instead see it as a pixel-to-pixel mapping of the guide image to the domain of the source image. The pixel-wise mapping is parameterised as a multi-layer perceptron, whose weights are learned by minimising the discrepancies between the source image and the downsampled target image. Importantly, our formulation makes it possible to regularise only the mapping function, while avoiding regularisation of the outputs; Thus producing crisp, natural-looking images. The proposed method is unsupervised, using only the specific source and guide images to fit the mapping. We evaluate our method on two different tasks, super-resolution of depth maps and of tree height maps. In both cases we clearly outperform recent baselines in quantitative comparisons, while delivering visually much sharper outputs.
http://arxiv.org/abs/1904.01501
Distributed word vector spaces are considered hard to interpret which hinders the understanding of natural language processing (NLP) models. In this work, we introduce a new method to interpret arbitrary samples from a word vector space. To this end, we train a neural model to conceptualize word vectors, which means that it activates higher order concepts it recognizes in a given vector. Contrary to prior approaches, our model operates in the original vector space and is capable of learning non-linear relations between word vectors and concepts. Furthermore, we show that it produces considerably less entropic concept activation profiles than the popular cosine similarity.
http://arxiv.org/abs/1904.01500
Real-world semantic or knowledge-based systems, e.g., in the biomedical domain, can become large and complex. Tool support for the localization and repair of faults within knowledge bases of such systems can therefore be essential for their practical success. Correspondingly, a number of knowledge base debugging approaches, in particular for ontology-based systems, were proposed throughout recent years. Query-based debugging is a comparably recent interactive approach that localizes the true cause of an observed problem by asking knowledge engineers a series of questions. Concrete implementations of this approach exist, such as the OntoDebug plug-in for the ontology editor Prot'eg'e. To validate that a newly proposed method is favorable over an existing one, researchers often rely on simulation-based comparisons. Such an evaluation approach however has certain limitations and often cannot fully inform us about a method’s true usefulness. We therefore conducted different user studies to assess the practical value of query-based ontology debugging. One main insight from the studies is that the considered interactive approach is indeed more efficient than an alternative algorithmic debugging based on test cases. We also observed that users frequently made errors in the process, which highlights the importance of a careful design of the queries that users need to answer.
http://arxiv.org/abs/1904.01484
Synthesizing photo-realistic images from text descriptions is a challenging problem. Previous studies have shown remarkable progresses on visual quality of the generated images. In this paper, we consider semantics from the input text descriptions in helping render photo-realistic images. However, diverse linguistic expressions pose challenges in extracting consistent semantics even they depict the same thing. To this end, we propose a novel photo-realistic text-to-image generation model that implicitly disentangles semantics to both fulfill the high-level semantic consistency and low-level semantic diversity. To be specific, we design (1) a Siamese mechanism in the discriminator to learn consistent high-level semantics, and (2) a visual-semantic embedding strategy by semantic-conditioned batch normalization to find diverse low-level semantics. Extensive experiments and ablation studies on CUB and MS-COCO datasets demonstrate the superiority of the proposed method in comparison to state-of-the-art methods.
http://arxiv.org/abs/1904.01480
Current image captioning systems perform at a merely descriptive level, essentially enumerating the objects in the scene and their relations. Humans, on the contrary, interpret images by integrating several sources of prior knowledge of the world. In this work, we aim to take a step closer to producing captions that offer a plausible interpretation of the scene, by integrating such contextual information into the captioning pipeline. For this we focus on the captioning of images used to illustrate news articles. We propose a novel captioning method that is able to leverage contextual information provided by the text of news articles associated with an image. Our model is able to selectively draw information from the article guided by visual cues, and to dynamically extend the output dictionary to out-of-vocabulary named entities that appear in the context source. Furthermore we introduce `GoodNews’, the largest news image captioning dataset in the literature and demonstrate state-of-the-art results.
http://arxiv.org/abs/1904.01475
Lemmatization aims to reduce the sparse data problem by relating the inflected forms of a word to its dictionary form. Using context can help, both for unseen and ambiguous words. Yet most context-sensitive approaches require full lemma-annotated sentences for training, which may be scarce or unavailable in low-resource languages. In addition (as shown here), in a low-resource setting, a lemmatizer can learn more from $n$ labeled examples of distinct words (types) than from $n$ (contiguous) labeled tokens, since the latter contain far fewer distinct types. To combine the efficiency of type-based learning with the benefits of context, we propose a way to train a context-sensitive lemmatizer with little or no labeled corpus data, using inflection tables from the UniMorph project and raw text examples from Wikipedia that provide sentence contexts for the unambiguous UniMorph examples. Despite these being unambiguous examples, the model successfully generalizes from them, leading to improved results (both overall, and especially on unseen words) in comparison to a baseline that does not use context.
http://arxiv.org/abs/1904.01464
Popular word embedding methods such as word2vec and GloVe assign a single vector representation to each word, even if a word has multiple distinct meanings. Multi-sense embeddings instead provide different vectors for each sense of a word. However, they typically cannot serve as a drop-in replacement for conventional single-sense embeddings, because the correct sense vector needs to be selected for each word. In this work, we study the effect of multi-sense embeddings on the task of reverse dictionaries. We propose a technique to easily integrate them into an existing neural network architecture using an attention mechanism. Our experiments demonstrate that large improvements can be obtained when employing multi-sense embeddings both in the input sequence as well as for the target representation. An analysis of the sense distributions and of the learned attention is provided as well.
http://arxiv.org/abs/1904.01451
Point set registration is defined as a process to determine the spatial transformation from the source point set to the target one. Existing methods often iteratively search for the optimal geometric transformation to register a given pair of point sets, driven by minimizing a predefined alignment loss function. In contrast, the proposed point registration neural network (PR-Net) actively learns the registration pattern as a parametric function from a training dataset, consequently predict the desired geometric transformation to align a pair of point sets. PR-Net can transfer the learned knowledge (i.e. registration pattern) from registering training pairs to testing ones without additional iterative optimization. Specifically, in this paper, we develop novel techniques to learn shape descriptors from point sets that help formulate a clear correlation between source and target point sets. With the defined correlation, PR-Net tends to predict the transformation so that the source and target point sets can be statistically aligned, which in turn leads to an optimal spatial geometric registration. PR-Net achieves robust and superior performance for non-rigid registration of point sets, even in presence of Gaussian noise, outliers, and missing points, but requires much less time for registering large number of pairs. More importantly, for a new pair of point sets, PR-Net is able to directly predict the desired transformation using the learned model without repetitive iterative optimization routine. Our code is available at https://github.com/Lingjing324/PR-Net.
http://arxiv.org/abs/1904.01428
The goal of this paper is to retrieve an image based on instance, attribute and category similarity notions. Different from existing works, which usually address only one of these entities in isolation, we introduce a cooperative embedding to integrate them while preserving their specific level of semantic representation. An algebraic structure defines a superspace filled with instances. Attributes are axis-aligned to form subspaces, while categories influence the arrangement of similar instances. These relationships enable them to cooperate for their mutual benefits for image retrieval. We derive a proxy-based softmax embedding loss to learn simultaneously all similarity measures in both superspace and subspaces. We evaluate our model on datasets from two different domains. Experiments on image retrieval tasks show the benefits of the cooperative embeddings for modeling multiple image similarities, and for discovering style evolution of instances between- and within-categories.
http://arxiv.org/abs/1904.01421
Semantic scene understanding is important for various applications. In particular, self-driving cars need a fine-grained understanding of the surfaces and objects in their vicinity. Light detection and ranging (LiDAR) provides precise geometric information about the environment and is thus a part of the sensor suites of almost all self-driving cars. Despite the relevance of semantic scene understanding for this application, there is a lack of a large dataset for this task which is based on an automotive LiDAR. In this paper, we introduce a large dataset to propel research on laser-based semantic segmentation. We annotated all sequences of the KITTI Vision Odometry Benchmark and provide dense point-wise annotations for the complete $360^{o}$ field-of-view of the employed automotive LiDAR. We propose three benchmark tasks based on this dataset: (i) semantic segmentation of point clouds using a single scan, (ii) semantic segmentation using sequences comprised of multiple past scans, and (iii) semantic scene completion, which requires to anticipate the semantic scene in the future. We provide baseline experiments and show that there is a need for more sophisticated models to efficiently tackle these tasks. Our dataset opens the door for the development of more advanced methods, but also provides plentiful data to investigate new research directions.
http://arxiv.org/abs/1904.01416
Dense captioning aims at simultaneously localizing semantic regions and describing these regions-of-interest (ROIs) with short phrases or sentences in natural language. Previous studies have shown remarkable progresses, but they are often vulnerable to the aperture problem that a caption generated by the features inside one ROI lacks contextual coherence with its surrounding context in the input image. In this work, we investigate contextual reasoning based on multi-scale message propagations from the neighboring contents to the target ROIs. To this end, we design a novel end-to-end context and attribute grounded dense captioning framework consisting of 1) a contextual visual mining module and 2) a multi-level attribute grounded description generation module. Knowing that captions often co-occur with the linguistic attributes (such as who, what and where), we also incorporate an auxiliary supervision from hierarchical linguistic attributes to augment the distinctiveness of the learned captions. Extensive experiments and ablation studies on Visual Genome dataset demonstrate the superiority of the proposed model in comparison to state-of-the-art methods.
http://arxiv.org/abs/1904.01410
In this work, we construct a large-scale dataset for vehicle re-identification (ReID), which contains 137k images of 13k vehicle instances captured by UAV-mounted cameras. To our knowledge, it is the largest UAV-based vehicle ReID dataset. To increase intra-class variation, each vehicle is captured by at least two UAVs at different locations, with diverse view-angles and flight-altitudes. We manually label a variety of vehicle attributes, including vehicle type, color, skylight, bumper, spare tire and luggage rack. Furthermore, for each vehicle image, the annotator is also required to mark the discriminative parts that helps them to distinguish this particular vehicle from others. Besides the dataset, we also design a specific vehicle ReID algorithm to make full use of the rich annotation information. It is capable of explicitly detecting discriminative parts for each specific vehicle and significantly outperforms the evaluated baselines and state-of-the-art vehicle ReID approaches.
http://arxiv.org/abs/1904.01400
Recognition of defects in concrete infrastructure, especially in bridges, is a costly and time consuming crucial first step in the assessment of the structural integrity. Large variation in appearance of the concrete material, changing illumination and weather conditions, a variety of possible surface markings as well as the possibility for different types of defects to overlap, make it a challenging real-world task. In this work we introduce the novel COncrete DEfect BRidge IMage dataset (CODEBRIM) for multi-target classification of five commonly appearing concrete defects. We investigate and compare two reinforcement learning based meta-learning approaches, MetaQNN and efficient neural architecture search, to find suitable convolutional neural network architectures for this challenging multi-class multi-target task. We show that learned architectures have fewer overall parameters in addition to yielding better multi-target accuracy in comparison to popular neural architectures from the literature evaluated in the context of our application.
http://arxiv.org/abs/1904.08486
We propose an effective deep learning approach to aesthetics quality assessment that relies on a new type of pre-trained features, and apply it to the AVA data set, the currently largest aesthetics database. While previous approaches miss some of the information in the original images, due to taking small crops, down-scaling or warping the originals during training, we propose the first method that efficiently supports full resolution images as an input, and can be trained on variable input sizes. This allows us to significantly improve upon the state of the art, increasing the Spearman rank-order correlation coefficient (SRCC) of ground-truth mean opinion scores (MOS) from the existing best reported of 0.612 to 0.756. To achieve this performance, we extract multi-level spatially pooled (MLSP) features from all convolutional blocks of a pre-trained InceptionResNet-v2 network, and train a custom shallow Convolutional Neural Network (CNN) architecture on these new features.
http://arxiv.org/abs/1904.01382
We propose 3DSmoothNet, a full workflow to match 3D point clouds with a siamese deep learning architecture and fully convolutional layers using a voxelized smoothed density value (SDV) representation. The latter is computed per interest point and aligned to the local reference frame (LRF) to achieve rotation invariance. Our compact, learned, rotation invariant 3D point cloud descriptor achieves 94.9% average recall on the 3DMatch benchmark data set, outperforming the state-of-the-art by more than 20 percent points with only 32 output dimensions. This very low output dimension allows for near realtime correspondence search with 0.1 ms per feature point on a standard PC. Our approach is sensor- and sceneagnostic because of SDV, LRF and learning highly descriptive features with fully convolutional layers. We show that 3DSmoothNet trained only on RGB-D indoor scenes of buildings achieves 79.0% average recall on laser scans of outdoor vegetation, more than double the performance of our closest, learning-based competitors. Code, data and pre-trained models are available online at this https URL.
https://arxiv.org/abs/1811.06879
Transfer learning aims at transferring knowledge from a well-labeled domain to a similar but different domain with limited or no labels. Unfortunately, existing learning-based methods often involve intensive model selection and hyperparameter tuning to obtain good results. Moreover, cross-validation is not possible for tuning hyperparameters since there are often no labels in the target domain. This would restrict wide applicability of transfer learning especially in computationally-constraint devices such as wearables. In this paper, we propose a practically Easy Transfer Learning (EasyTL) approach which requires no model selection and hyperparameter tuning, while achieving competitive performance. By exploiting intra-domain structures, EasyTL is able to learn both non-parametric transfer features and classifiers. Extensive experiments demonstrate that, compared to state-of-the-art traditional and deep methods, EasyTL satisfies the Occam’s Razor principle: it is extremely easy to implement and use while achieving comparable or better performance in classification accuracy and much better computational efficiency. Additionally, it is shown that EasyTL can increase the performance of existing transfer feature learning methods.
http://arxiv.org/abs/1904.01376
Reading irregular text of arbitrary shape in natural scene images is still a challenging problem. Many existing approaches incorporate sophisticated network structures to handle various shapes, use extra annotations for stronger supervision, or employ hard-to-train recurrent neural networks for sequence modeling. In this work, we propose a simple yet robust approach for irregular text recognition. With no need to convert input images to sequence representations, we directly connect two-dimensional CNN features to an attention-based sequence decoder. As no recurrent module is adopted, our model can be trained in parallel. It achieves 3x to 18x acceleration to backward pass and 2x to 12x acceleration to forward pass, compared with the RNN counterparts. The proposed model is trained with only word-level annotations. With this simple design, our method achieves state-of-the-art or competitive recognition performance on the evaluated regular and irregular scene text benchmark datasets. Furthermore, we show that the recognition performance does not significantly degrade with inaccurate bounding boxes. This is desirable for tasks of end-to-end text detection and recognition: robust recognition performance can still be achieved with an inaccurate text detector. We will release the code.
http://arxiv.org/abs/1904.01375
Edge-preserving image smoothing is an important step for many low-level vision problems. Though many algorithms have been proposed, there are several difficulties hindering its further development. First, most existing algorithms cannot perform well on a wide range of image contents using a single parameter setting. Second, the performance evaluation of edge-preserving image smoothing remains subjective, and there lacks a widely accepted datasets to objectively compare the different algorithms. To address these issues and further advance the state of the art, in this work we propose a benchmark for edge-preserving image smoothing. This benchmark includes an image dataset with groundtruth image smoothing results as well as baseline algorithms that can generate competitive edge-preserving smoothing results for a wide range of image contents. The established dataset contains 500 training and testing images with a number of representative visual object categories, while the baseline methods in our benchmark are built upon representative deep convolutional network architectures, on top of which we design novel loss functions well suited for edge-preserving image smoothing. The trained deep networks run faster than most state-of-the-art smoothing algorithms with leading smoothing results both qualitatively and quantitatively. The benchmark is publicly accessible via this https URL.
https://arxiv.org/abs/1904.01579
We present an unsupervised training approach for a neural network-based mask estimator in an acoustic beamforming application. The network is trained to maximize a likelihood criterion derived from a spatial mixture model of the observations. It is trained from scratch without requiring any parallel data consisting of degraded input and clean training targets. Thus, training can be carried out on real recordings of noisy speech rather than simulated ones. In contrast to previous work on unsupervised training of neural mask estimators, our approach avoids the need for a possibly pre-trained teacher model entirely. We demonstrate the effectiveness of our approach by speech recognition experiments on two different datasets: one mainly deteriorated by noise (CHiME 4) and one by reverberation (REVERB). The results show that the performance of the proposed system is on par with a supervised system using oracle target masks for training and with a system trained using a model-based teacher.
https://arxiv.org/abs/1904.01578
Photon-limited images are often seen in fields such as medical imaging. Although the number of collected photons on an image sensor statistically follows Poisson distribution, this type of noise is intractable, unlike Gaussian noise. In this study, we propose a Bayesian restoration method of Poisson corrupted image using Integrated Nested Laplace Approximation (INLA), which is a computational method to evaluate marginalized posterior distributions of latent Gaussian models (LGMs). When the original image can be regarded as ICAR (intrinsic conditional auto-regressive) model reasonably, our method performs very faster than well-known ones such as loopy belief propagation-based method and Markov chain Monte Carlo (MCMC) without decreasing the accuracy.
http://arxiv.org/abs/1904.01357
With massive explosion of social media such as Twitter and Instagram, people daily share billions of multimedia posts, containing images and text. Typically, text in these posts is short, informal and noisy, leading to ambiguities which can be resolved using images. In this paper we explore text-centric Named Entity Recognition task on these multimedia posts. We propose an end to end model which learns a joint representation of a text and an image. Our model extends multi-dimensional self attention technique, where now image helps to enhance relationship between words. Experiments show that our model is capable of capturing both textual and visual contexts with greater accuracy, achieving state-of-the-art results on Twitter multimodal Named Entity Recognition dataset.
http://arxiv.org/abs/1904.01356
We propose a fully convolutional one-stage object detector (FCOS) to solve object detection in a per-pixel prediction fashion, analogue to semantic segmentation. Almost all state-of-the-art object detectors such as RetinaNet, SSD, YOLOv3, and Faster R-CNN rely on pre-defined anchor boxes. In contrast, our proposed detector FCOS is anchor-box free, as well as proposal free. By eliminating the pre-defined set of anchor boxes, FCOS completely avoids the complicated computation related to anchor boxes such as calculating overlapping during training and significantly reduces the training memory footprint. More importantly, we also avoid all hyper-parameters related to anchor boxes, which are often very sensitive to the final detection performance. With the only post-processing non-maximum suppression (NMS), our detector FCOS outperforms previous anchor-based one-stage detectors with the advantage of being much simpler. For the first time, we demonstrate a much simpler and flexible detection framework achieving improved detection accuracy. We hope that the proposed FCOS framework can serve as a simple and strong alternative for many other instance-level tasks.
http://arxiv.org/abs/1904.01355
In this paper, we solve the problem of adapting classifiers across domains. We consider the problem of domain adaptation for multi-class classification where we are provided a labeled set of examples in a source dataset and we are provided a target dataset with no supervision. In this setting, we propose an adversarial discriminator based approach. While the approach based on adversarial discriminator has been previously proposed; in this paper, we present an informed adversarial discriminator. Our observation relies on the analysis that shows that if the discriminator has access to all the information available including the class structure present in the source dataset, then it can guide the transformation of features of the target set of classes to a more structure adapted space. Using this formulation, we obtain state-of-the-art results for the standard evaluation on benchmark datasets. We further provide detailed analysis which shows that using all the labeled information results in an improved domain adaptation.
http://arxiv.org/abs/1904.01341
In this article a novel approach for training deep neural networks using Bayesian techniques is presented. The Bayesian methodology allows for an easy evaluation of model uncertainty and additionally is robust to overfitting. These are commonly the two main problems classical, i.e. non-Bayesian, architectures have to struggle with. The proposed approach applies variational inference in order to approximate the intractable posterior distribution. In particular, the variational distribution is defined as product of multiple multivariate normal distributions with tridiagonal covariance matrices. Each single normal distribution belongs either to the weights, or to the biases corresponding to one network layer. The layer-wise a posteriori variances are defined based on the corresponding expectation values and further the correlations are assumed to be identical. Therefore, only a few additional parameters need to be optimized compared to non-Bayesian settings. The novel approach is successfully evaluated on basis of the popular benchmark datasets MNIST and CIFAR-10.
http://arxiv.org/abs/1904.01334