Paper Group ANR 306
Fully Convolutional Network for Automatic Road Extraction from Satellite Imagery. Forward Stability of ResNet and Its Variants. Automatic Pixelwise Object Labeling for Aerial Imagery Using Stacked U-Nets. Product Characterisation towards Personalisation: Learning Attributes from Unstructured Data to Recommend Fashion Products. MimicGAN: Corruption- …
Fully Convolutional Network for Automatic Road Extraction from Satellite Imagery
| Title | Fully Convolutional Network for Automatic Road Extraction from Satellite Imagery |
| Authors | Alexander V. Buslaev, Selim S. Seferbekov, Vladimir I. Iglovikov, Alexey A. Shvets |
| Abstract | Analysis of high-resolution satellite images has been an important research topic for traffic management, city planning, and road monitoring. One of the problems here is automatic and precise road extraction. From an original image, it is difficult and computationally expensive to extract roads due to presences of other road-like features with straight edges. In this paper, we propose an approach for automatic road extraction based on a fully convolutional neural network of U-net family. This network consists of ResNet-34 pre-trained on ImageNet and decoder adapted from vanilla U-Net. Based on validation results, leaderboard and our own experience this network shows superior results for the DEEPGLOBE - CVPR 2018 road extraction sub-challenge. Moreover, this network uses moderate memory that allows using just one GTX 1080 or 1080ti video cards to perform whole training and makes pretty fast predictions. |
| Tasks | |
| Published | 2018-06-13 |
| URL | http://arxiv.org/abs/1806.05182v2 |
| http://arxiv.org/pdf/1806.05182v2.pdf | |
| PWC | https://paperswithcode.com/paper/fully-convolutional-network-for-automatic |
| Repo | |
| Framework | |
Forward Stability of ResNet and Its Variants
| Title | Forward Stability of ResNet and Its Variants |
| Authors | Linan Zhang, Hayden Schaeffer |
| Abstract | The residual neural network (ResNet) is a popular deep network architecture which has the ability to obtain high-accuracy results on several image processing problems. In order to analyze the behavior and structure of ResNet, recent work has been on establishing connections between ResNets and continuous-time optimal control problems. In this work, we show that the post-activation ResNet is related to an optimal control problem with differential inclusions, and provide continuous-time stability results for the differential inclusion associated with ResNet. Motivated by the stability conditions, we show that alterations of either the architecture or the optimization problem can generate variants of ResNet which improve the theoretical stability bounds. In addition, we establish stability bounds for the full (discrete) network associated with two variants of ResNet, in particular, bounds on the growth of the features and a measure of the sensitivity of the features with respect to perturbations. These results also help to show the relationship between the depth, regularization, and stability of the feature space. Computational experiments on the proposed variants show that the accuracy of ResNet is preserved and that the accuracy seems to be monotone with respect to the depth and various corruptions. |
| Tasks | |
| Published | 2018-11-24 |
| URL | http://arxiv.org/abs/1811.09885v1 |
| http://arxiv.org/pdf/1811.09885v1.pdf | |
| PWC | https://paperswithcode.com/paper/forward-stability-of-resnet-and-its-variants |
| Repo | |
| Framework | |
Automatic Pixelwise Object Labeling for Aerial Imagery Using Stacked U-Nets
| Title | Automatic Pixelwise Object Labeling for Aerial Imagery Using Stacked U-Nets |
| Authors | Andrew Khalel, Motaz El-Saban |
| Abstract | Automation of objects labeling in aerial imagery is a computer vision task with numerous practical applications. Fields like energy exploration require an automated method to process a continuous stream of imagery on a daily basis. In this paper we propose a pipeline to tackle this problem using a stack of convolutional neural networks (U-Net architecture) arranged end-to-end. Each network works as post-processor to the previous one. Our model outperforms current state-of-the-art on two different datasets: Inria Aerial Image Labeling dataset and Massachusetts Buildings dataset each with different characteristics such as spatial resolution, object shapes and scales. Moreover, we experimentally validate computation time savings by processing sub-sampled images and later upsampling pixelwise labeling. These savings come at a negligible degradation in segmentation quality. Though the conducted experiments in this paper cover only aerial imagery, the technique presented is general and can handle other types of images. |
| Tasks | |
| Published | 2018-03-13 |
| URL | http://arxiv.org/abs/1803.04953v1 |
| http://arxiv.org/pdf/1803.04953v1.pdf | |
| PWC | https://paperswithcode.com/paper/automatic-pixelwise-object-labeling-for |
| Repo | |
| Framework | |
Product Characterisation towards Personalisation: Learning Attributes from Unstructured Data to Recommend Fashion Products
| Title | Product Characterisation towards Personalisation: Learning Attributes from Unstructured Data to Recommend Fashion Products |
| Authors | Ângelo Cardoso, Fabio Daolio, Saúl Vargas |
| Abstract | In this paper, we describe a solution to tackle a common set of challenges in e-commerce, which arise from the fact that new products are continually being added to the catalogue. The challenges involve properly personalising the customer experience, forecasting demand and planning the product range. We argue that the foundational piece to solve all of these problems is having consistent and detailed information about each product, information that is rarely available or consistent given the multitude of suppliers and types of products. We describe in detail the architecture and methodology implemented at ASOS, one of the world’s largest fashion e-commerce retailers, to tackle this problem. We then show how this quantitative understanding of the products can be leveraged to improve recommendations in a hybrid recommender system approach. |
| Tasks | Recommendation Systems |
| Published | 2018-03-20 |
| URL | http://arxiv.org/abs/1803.07679v1 |
| http://arxiv.org/pdf/1803.07679v1.pdf | |
| PWC | https://paperswithcode.com/paper/product-characterisation-towards |
| Repo | |
| Framework | |
MimicGAN: Corruption-Mimicking for Blind Image Recovery & Adversarial Defense
| Title | MimicGAN: Corruption-Mimicking for Blind Image Recovery & Adversarial Defense |
| Authors | Rushil Anirudh, Jayaraman J. Thiagarajan, Bhavya Kailkhura, Timo Bremer |
| Abstract | Solving inverse problems continues to be a central challenge in computer vision. Existing techniques either explicitly construct an inverse mapping using prior knowledge about the corruption, or learn the inverse directly using a large collection of examples. However, in practice, the nature of corruption may be unknown, and thus it is challenging to regularize the problem of inferring a plausible solution. On the other hand, collecting task-specific training data is tedious for known corruptions and impossible for unknown ones. We present MimicGAN, an unsupervised technique to solve general inverse problems based on image priors in the form of generative adversarial networks (GANs). Using a GAN prior, we show that one can reliably recover solutions to underdetermined inverse problems through a surrogate network that learns to mimic the corruption at test time. Our system successively estimates the corruption and the clean image without the need for supervisory training, while outperforming existing baselines in blind image recovery. We also demonstrate that MimicGAN improves upon recent GAN-based defenses against adversarial attacks and represents one of the strongest test-time defenses available today. |
| Tasks | Adversarial Defense |
| Published | 2018-11-20 |
| URL | http://arxiv.org/abs/1811.08484v1 |
| http://arxiv.org/pdf/1811.08484v1.pdf | |
| PWC | https://paperswithcode.com/paper/mimicgan-corruption-mimicking-for-blind-image |
| Repo | |
| Framework | |
MSDNN: Multi-Scale Deep Neural Network for Salient Object Detection
| Title | MSDNN: Multi-Scale Deep Neural Network for Salient Object Detection |
| Authors | Fen Xiao, Wenzheng Deng, Liangchan Peng, Chunhong Cao, Kai Hu, Xieping Gao |
| Abstract | Salient object detection is a fundamental problem and has been received a great deal of attentions in computer vision. Recently deep learning model became a powerful tool for image feature extraction. In this paper, we propose a multi-scale deep neural network (MSDNN) for salient object detection. The proposed model first extracts global high-level features and context information over the whole source image with recurrent convolutional neural network (RCNN). Then several stacked deconvolutional layers are adopted to get the multi-scale feature representation and obtain a series of saliency maps. Finally, we investigate a fusion convolution module (FCM) to build a final pixel level saliency map. The proposed model is extensively evaluated on four salient object detection benchmark datasets. Results show that our deep model significantly outperforms other 12 state-of-the-art approaches. |
| Tasks | Object Detection, Salient Object Detection |
| Published | 2018-01-12 |
| URL | http://arxiv.org/abs/1801.04187v1 |
| http://arxiv.org/pdf/1801.04187v1.pdf | |
| PWC | https://paperswithcode.com/paper/msdnn-multi-scale-deep-neural-network-for |
| Repo | |
| Framework | |
Combating catastrophic forgetting with developmental compression
| Title | Combating catastrophic forgetting with developmental compression |
| Authors | Shawn L. E. Beaulieu, Sam Kriegman, Josh C. Bongard |
| Abstract | Generally intelligent agents exhibit successful behavior across problems in several settings. Endemic in approaches to realize such intelligence in machines is catastrophic forgetting: sequential learning corrupts knowledge obtained earlier in the sequence, or tasks antagonistically compete for system resources. Methods for obviating catastrophic forgetting have sought to identify and preserve features of the system necessary to solve one problem when learning to solve another, or to enforce modularity such that minimally overlapping sub-functions contain task specific knowledge. While successful, both approaches scale poorly because they require larger architectures as the number of training instances grows, causing different parts of the system to specialize for separate subsets of the data. Here we present a method for addressing catastrophic forgetting called developmental compression. It exploits the mild impacts of developmental mutations to lessen adverse changes to previously-evolved capabilities and `compresses’ specialized neural networks into a generalized one. In the absence of domain knowledge, developmental compression produces systems that avoid overt specialization, alleviating the need to engineer a bespoke system for every task permutation and suggesting better scalability than existing approaches. We validate this method on a robot control problem and hope to extend this approach to other machine learning domains in the future. | |
| Tasks | |
| Published | 2018-04-12 |
| URL | http://arxiv.org/abs/1804.04286v1 |
| http://arxiv.org/pdf/1804.04286v1.pdf | |
| PWC | https://paperswithcode.com/paper/combating-catastrophic-forgetting-with |
| Repo | |
| Framework | |
Online Aggregation of Unbounded Losses Using Shifting Experts with Confidence
| Title | Online Aggregation of Unbounded Losses Using Shifting Experts with Confidence |
| Authors | Vladimir V’yugin, Vladimir Trunov |
| Abstract | We develop the setting of sequential prediction based on shifting experts and on a “smooth” version of the method of specialized experts. To aggregate experts predictions, we use the AdaHedge algorithm, which is a version of the Hedge algorithm with adaptive learning rate, and extend it by the meta-algorithm Fixed Share. Due to this, we combine the advantages of both algorithms: (1) we use the shifting regret which is a more optimal characteristic of the algorithm; (2) regret bounds are valid in the case of signed unbounded losses of the experts. Also, (3) we incorporate in this scheme a “smooth” version of the method of specialized experts which allows us to make more flexible and accurate predictions. All results are obtained in the adversarial setting – no assumptions are made about the nature of data source. We present results of numerical experiments for short-term forecasting of electricity consumption based on a real data. |
| Tasks | |
| Published | 2018-08-02 |
| URL | https://arxiv.org/abs/1808.00741v3 |
| https://arxiv.org/pdf/1808.00741v3.pdf | |
| PWC | https://paperswithcode.com/paper/online-aggregation-of-unbounded-losses-using |
| Repo | |
| Framework | |
Transductive Unbiased Embedding for Zero-Shot Learning
| Title | Transductive Unbiased Embedding for Zero-Shot Learning |
| Authors | Jie Song, Chengchao Shen, Yezhou Yang, Yang Liu, Mingli Song |
| Abstract | Most existing Zero-Shot Learning (ZSL) methods have the strong bias problem, in which instances of unseen (target) classes tend to be categorized as one of the seen (source) classes. So they yield poor performance after being deployed in the generalized ZSL settings. In this paper, we propose a straightforward yet effective method named Quasi-Fully Supervised Learning (QFSL) to alleviate the bias problem. Our method follows the way of transductive learning, which assumes that both the labeled source images and unlabeled target images are available for training. In the semantic embedding space, the labeled source images are mapped to several fixed points specified by the source categories, and the unlabeled target images are forced to be mapped to other points specified by the target categories. Experiments conducted on AwA2, CUB and SUN datasets demonstrate that our method outperforms existing state-of-the-art approaches by a huge margin of 9.3~24.5% following generalized ZSL settings, and by a large margin of 0.2~16.2% following conventional ZSL settings. |
| Tasks | Zero-Shot Learning |
| Published | 2018-03-30 |
| URL | http://arxiv.org/abs/1803.11320v1 |
| http://arxiv.org/pdf/1803.11320v1.pdf | |
| PWC | https://paperswithcode.com/paper/transductive-unbiased-embedding-for-zero-shot |
| Repo | |
| Framework | |
Human Perceptions of Fairness in Algorithmic Decision Making: A Case Study of Criminal Risk Prediction
| Title | Human Perceptions of Fairness in Algorithmic Decision Making: A Case Study of Criminal Risk Prediction |
| Authors | Nina Grgić-Hlača, Elissa M. Redmiles, Krishna P. Gummadi, Adrian Weller |
| Abstract | As algorithms are increasingly used to make important decisions that affect human lives, ranging from social benefit assignment to predicting risk of criminal recidivism, concerns have been raised about the fairness of algorithmic decision making. Most prior works on algorithmic fairness normatively prescribe how fair decisions ought to be made. In contrast, here, we descriptively survey users for how they perceive and reason about fairness in algorithmic decision making. A key contribution of this work is the framework we propose to understand why people perceive certain features as fair or unfair to be used in algorithms. Our framework identifies eight properties of features, such as relevance, volitionality and reliability, as latent considerations that inform people’s moral judgments about the fairness of feature use in decision-making algorithms. We validate our framework through a series of scenario-based surveys with 576 people. We find that, based on a person’s assessment of the eight latent properties of a feature in our exemplar scenario, we can accurately (> 85%) predict if the person will judge the use of the feature as fair. Our findings have important implications. At a high-level, we show that people’s unfairness concerns are multi-dimensional and argue that future studies need to address unfairness concerns beyond discrimination. At a low-level, we find considerable disagreements in people’s fairness judgments. We identify root causes of the disagreements, and note possible pathways to resolve them. |
| Tasks | Decision Making |
| Published | 2018-02-26 |
| URL | http://arxiv.org/abs/1802.09548v1 |
| http://arxiv.org/pdf/1802.09548v1.pdf | |
| PWC | https://paperswithcode.com/paper/human-perceptions-of-fairness-in-algorithmic |
| Repo | |
| Framework | |
Towards Arbitrary Noise Augmentation - Deep Learning for Sampling from Arbitrary Probability Distributions
| Title | Towards Arbitrary Noise Augmentation - Deep Learning for Sampling from Arbitrary Probability Distributions |
| Authors | Felix Horger, Tobias Würfl, Vincent Christlein, Andreas Maier |
| Abstract | Accurate noise modelling is important for training of deep learning reconstruction algorithms. While noise models are well known for traditional imaging techniques, the noise distribution of a novel sensor may be difficult to determine a priori. Therefore, we propose learning arbitrary noise distributions. To do so, this paper proposes a fully connected neural network model to map samples from a uniform distribution to samples of any explicitly known probability density function. During the training, the Jensen-Shannon divergence between the distribution of the model’s output and the target distribution is minimized. We experimentally demonstrate that our model converges towards the desired state. It provides an alternative to existing sampling methods such as inversion sampling, rejection sampling, Gaussian mixture models and Markov-Chain-Monte-Carlo. Our model has high sampling efficiency and is easily applied to any probability distribution, without the need of further analytical or numerical calculations. |
| Tasks | |
| Published | 2018-01-12 |
| URL | http://arxiv.org/abs/1801.04211v2 |
| http://arxiv.org/pdf/1801.04211v2.pdf | |
| PWC | https://paperswithcode.com/paper/towards-arbitrary-noise-augmentation-deep |
| Repo | |
| Framework | |
An Investigation of the Interactions Between Pre-Trained Word Embeddings, Character Models and POS Tags in Dependency Parsing
| Title | An Investigation of the Interactions Between Pre-Trained Word Embeddings, Character Models and POS Tags in Dependency Parsing |
| Authors | Aaron Smith, Miryam de Lhoneux, Sara Stymne, Joakim Nivre |
| Abstract | We provide a comprehensive analysis of the interactions between pre-trained word embeddings, character models and POS tags in a transition-based dependency parser. While previous studies have shown POS information to be less important in the presence of character models, we show that in fact there are complex interactions between all three techniques. In isolation each produces large improvements over a baseline system using randomly initialised word embeddings only, but combining them quickly leads to diminishing returns. We categorise words by frequency, POS tag and language in order to systematically investigate how each of the techniques affects parsing quality. For many word categories, applying any two of the three techniques is almost as good as the full combined system. Character models tend to be more important for low-frequency open-class words, especially in morphologically rich languages, while POS tags can help disambiguate high-frequency function words. We also show that large character embedding sizes help even for languages with small character sets, especially in morphologically rich languages. |
| Tasks | Dependency Parsing, Word Embeddings |
| Published | 2018-08-27 |
| URL | http://arxiv.org/abs/1808.09060v1 |
| http://arxiv.org/pdf/1808.09060v1.pdf | |
| PWC | https://paperswithcode.com/paper/an-investigation-of-the-interactions-between |
| Repo | |
| Framework | |
Spectral Simplicial Theory for Feature Selection and Applications to Genomics
| Title | Spectral Simplicial Theory for Feature Selection and Applications to Genomics |
| Authors | Kiya W. Govek, Venkata S. Yamajala, Pablo G. Camara |
| Abstract | The scale and complexity of modern data sets and the limitations associated with testing large numbers of hypotheses underline the need for feature selection methods. Spectral techniques rank features according to their degree of consistency with an underlying metric structure, but their current graph-based formulation restricts their applicability to point features. We extend spectral methods for feature selection to abstract simplicial complexes and present a general framework which can be applied to 2-point and higher-order features. Combinatorial Laplacian scores take into account the topology spanned by the data and reduce to the ordinary Laplacian score in the case of point features. We demonstrate the utility of spectral simplicial methods for feature selection with several examples of application to the analysis of gene expression and multi-modal genomic data. Our results provide a unifying perspective on topological data analysis and manifold learning approaches. |
| Tasks | Feature Selection, Topological Data Analysis |
| Published | 2018-11-08 |
| URL | http://arxiv.org/abs/1811.03377v1 |
| http://arxiv.org/pdf/1811.03377v1.pdf | |
| PWC | https://paperswithcode.com/paper/spectral-simplicial-theory-for-feature |
| Repo | |
| Framework | |
Precise but Natural Specification for Robot Tasks
| Title | Precise but Natural Specification for Robot Tasks |
| Authors | Ivan Gavran, Brendon Boldt, Eva Darulova, Rupak Majumdar |
| Abstract | We present Flipper, a natural language interface for describing high-level task specifications for robots that are compiled into robot actions. Flipper starts with a formal core language for task planning that allows expressing rich temporal specifications and uses a semantic parser to provide a natural language interface. Flipper provides immediate visual feedback by executing an automatically constructed plan of the task in a graphical user interface. This allows the user to resolve potentially ambiguous interpretations. Flipper extends itself via naturalization: its users can add definitions for utterances, from which Flipper induces new rules and adds them to the core language, gradually growing a more and more natural task specification language. Flipper improves the naturalization by generalizing the definition provided by users. Unlike other task-specification systems, Flipper enables natural language interactions while maintaining the expressive power and formal precision of a programming language. We show through an initial user study that natural language interactions and generalization can considerably ease the description of tasks. Moreover, over time, users employ more and more concepts outside of the initial core language. Such extensions are available to the Flipper community, and users can use concepts that others have defined. |
| Tasks | |
| Published | 2018-03-06 |
| URL | http://arxiv.org/abs/1803.02238v2 |
| http://arxiv.org/pdf/1803.02238v2.pdf | |
| PWC | https://paperswithcode.com/paper/precise-but-natural-specification-for-robot |
| Repo | |
| Framework | |
On stochastic gradient Langevin dynamics with dependent data streams in the logconcave case
| Title | On stochastic gradient Langevin dynamics with dependent data streams in the logconcave case |
| Authors | M. Barkhagen, N. H. Chau, É. Moulines, M. Rásonyi, S. Sabanis, Y. Zhang |
| Abstract | We study the problem of sampling from a probability distribution $\pi$ on $\rset^d$ which has a density \wrt\ the Lebesgue measure known up to a normalization factor $x \mapsto \rme^{-U(x)} / \int_{\rset^d} \rme^{-U(y)} \rmd y$. We analyze a sampling method based on the Euler discretization of the Langevin stochastic differential equations under the assumptions that the potential $U$ is continuously differentiable, $\nabla U$ is Lipschitz, and $U$ is strongly concave. We focus on the case where the gradient of the log-density cannot be directly computed but unbiased estimates of the gradient from possibly dependent observations are available. This setting can be seen as a combination of a stochastic approximation (here stochastic gradient) type algorithms with discretized Langevin dynamics. We obtain an upper bound of the Wasserstein-2 distance between the law of the iterates of this algorithm and the target distribution $\pi$ with constants depending explicitly on the Lipschitz and strong convexity constants of the potential and the dimension of the space. Finally, under weaker assumptions on $U$ and its gradient but in the presence of independent observations, we obtain analogous results in Wasserstein-2 distance. |
| Tasks | Stochastic Optimization |
| Published | 2018-12-06 |
| URL | https://arxiv.org/abs/1812.02709v3 |
| https://arxiv.org/pdf/1812.02709v3.pdf | |
| PWC | https://paperswithcode.com/paper/on-stochastic-gradient-langevin-dynamics-with |
| Repo | |
| Framework | |
