May 7, 2019

2755 words 13 mins read

Paper Group ANR 75

Optimizing Recurrent Neural Networks Architectures under Time Constraints. Strategic Attentive Writer for Learning Macro-Actions. Latent fingerprint minutia extraction using fully convolutional network. A Modified Vortex Search Algorithm for Numerical Function Optimization. Zero-error dissimilarity based classifiers. Detecção de comunidades em rede …

Optimizing Recurrent Neural Networks Architectures under Time Constraints


Title	Optimizing Recurrent Neural Networks Architectures under Time Constraints
Authors	Junqi Jin, Ziang Yan, Kun Fu, Nan Jiang, Changshui Zhang
Abstract	Recurrent neural network (RNN)‘s architecture is a key factor influencing its performance. We propose algorithms to optimize hidden sizes under running time constraint. We convert the discrete optimization into a subset selection problem. By novel transformations, the objective function becomes submodular and constraint becomes supermodular. A greedy algorithm with bounds is suggested to solve the transformed problem. And we show how transformations influence the bounds. To speed up optimization, surrogate functions are proposed which balance exploration and exploitation. Experiments show that our algorithms can find more accurate models or faster models than manually tuned state-of-the-art and random search. We also compare popular RNN architectures using our algorithms.
Tasks
Published	2016-08-29
URL	http://arxiv.org/abs/1608.07892v3
PDF	http://arxiv.org/pdf/1608.07892v3.pdf
PWC	https://paperswithcode.com/paper/optimizing-recurrent-neural-networks
Repo
Framework

Strategic Attentive Writer for Learning Macro-Actions


Title	Strategic Attentive Writer for Learning Macro-Actions
Authors	Alexander, Vezhnevets, Volodymyr Mnih, John Agapiou, Simon Osindero, Alex Graves, Oriol Vinyals, Koray Kavukcuoglu
Abstract	We present a novel deep recurrent neural network architecture that learns to build implicit plans in an end-to-end manner by purely interacting with an environment in reinforcement learning setting. The network builds an internal plan, which is continuously updated upon observation of the next input from the environment. It can also partition this internal representation into contiguous sub- sequences by learning for how long the plan can be committed to - i.e. followed without re-planing. Combining these properties, the proposed model, dubbed STRategic Attentive Writer (STRAW) can learn high-level, temporally abstracted macro- actions of varying lengths that are solely learnt from data without any prior information. These macro-actions enable both structured exploration and economic computation. We experimentally demonstrate that STRAW delivers strong improvements on several ATARI games by employing temporally extended planning strategies (e.g. Ms. Pacman and Frostbite). It is at the same time a general algorithm that can be applied on any sequence data. To that end, we also show that when trained on text prediction task, STRAW naturally predicts frequent n-grams (instead of macro-actions), demonstrating the generality of the approach.
Tasks	Atari Games
Published	2016-06-15
URL	http://arxiv.org/abs/1606.04695v1
PDF	http://arxiv.org/pdf/1606.04695v1.pdf
PWC	https://paperswithcode.com/paper/strategic-attentive-writer-for-learning-macro
Repo
Framework

Latent fingerprint minutia extraction using fully convolutional network


Title	Latent fingerprint minutia extraction using fully convolutional network
Authors	Yao Tang, Fei Gao, Jufu Feng
Abstract	Minutiae play a major role in fingerprint identification. Extracting reliable minutiae is difficult for latent fingerprints which are usually of poor quality. As the limitation of traditional handcrafted features, a fully convolutional network (FCN) is utilized to learn features directly from data to overcome complex background noises. Raw fingerprints are mapped to a correspondingly-sized minutia-score map with a fixed stride. And thus a large number of minutiae will be extracted through a given threshold. Then small regions centering at these minutia points are entered into a convolutional neural network (CNN) to reclassify these minutiae and calculate their orientations. The CNN shares convolutional layers with the fully convolutional network to speed up. 0.45 second is used on average to detect one fingerprint on a GPU. On the NIST SD27 database, we achieve 53% recall rate and 53% precise rate that outperform many other algorithms. Our trained model is also visualized to show that we have successfully extracted features preserving ridge information of a latent fingerprint.
Tasks
Published	2016-09-30
URL	http://arxiv.org/abs/1609.09850v2
PDF	http://arxiv.org/pdf/1609.09850v2.pdf
PWC	https://paperswithcode.com/paper/latent-fingerprint-minutia-extraction-using
Repo
Framework

A Modified Vortex Search Algorithm for Numerical Function Optimization


Title	A Modified Vortex Search Algorithm for Numerical Function Optimization
Authors	Berat Doğan
Abstract	The Vortex Search (VS) algorithm is one of the recently proposed metaheuristic algorithms which was inspired from the vortical flow of the stirred fluids. Although the VS algorithm is shown to be a good candidate for the solution of certain optimization problems, it also has some drawbacks. In the VS algorithm, candidate solutions are generated around the current best solution by using a Gaussian distribution at each iteration pass. This provides simplicity to the algorithm but it also leads to some problems along. Especially, for the functions those have a number of local minimum points, to select a single point to generate candidate solutions leads the algorithm to being trapped into a local minimum point. Due to the adaptive step-size adjustment scheme used in the VS algorithm, the locality of the created candidate solutions is increased at each iteration pass. Therefore, if the algorithm cannot escape a local point as quickly as possible, it becomes much more difficult for the algorithm to escape from that point in the latter iterations. In this study, a modified Vortex Search algorithm (MVS) is proposed to overcome above mentioned drawback of the existing VS algorithm. In the MVS algorithm, the candidate solutions are generated around a number of points at each iteration pass. Computational results showed that with the help of this modification the global search ability of the existing VS algorithm is improved and the MVS algorithm outperformed the existing VS algorithm, PSO2011 and ABC algorithms for the benchmark numerical function set.
Tasks
Published	2016-06-08
URL	http://arxiv.org/abs/1606.02710v1
PDF	http://arxiv.org/pdf/1606.02710v1.pdf
PWC	https://paperswithcode.com/paper/a-modified-vortex-search-algorithm-for
Repo
Framework

Zero-error dissimilarity based classifiers


Title	Zero-error dissimilarity based classifiers
Authors	Robert P. W. Duin, Elzbieta Pekalska
Abstract	We consider general non-Euclidean distance measures between real world objects that need to be classified. It is assumed that objects are represented by distances to other objects only. Conditions for zero-error dissimilarity based classifiers are derived. Additional conditions are given under which the zero-error decision boundary is a continues function of the distances to a finite set of training samples. These conditions affect the objects as well as the distance measure used. It is argued that they can be met in practice.
Tasks
Published	2016-01-18
URL	http://arxiv.org/abs/1601.04451v1
PDF	http://arxiv.org/pdf/1601.04451v1.pdf
PWC	https://paperswithcode.com/paper/zero-error-dissimilarity-based-classifiers
Repo
Framework

Detecção de comunidades em redes complexas para identificar gargalos e desperdício de recursos em sistemas de ônibus


Title	Detecção de comunidades em redes complexas para identificar gargalos e desperdício de recursos em sistemas de ônibus
Authors	Carlos Caminha, Vasco Furtado, Vládia Pinheiro, Caio Ponte
Abstract	We propose here a methodology to help to understand the shortcomings of public transportation in a city via the mining of complex networks representing the supply and demand of public transport. We show how to build these networks based upon data on smart card use in buses via the application of algorithms that estimate an OD and reconstruct the complete itinerary of the passengers. The overlapping of the two networks sheds light in potential overload and waste in the offer of resources that can be mitigated with strategies for balancing supply and demand.
Tasks
Published	2016-06-12
URL	http://arxiv.org/abs/1606.03737v2
PDF	http://arxiv.org/pdf/1606.03737v2.pdf
PWC	https://paperswithcode.com/paper/deteccao-de-comunidades-em-redes-complexas
Repo
Framework

Scene Grammars, Factor Graphs, and Belief Propagation


Title	Scene Grammars, Factor Graphs, and Belief Propagation
Authors	Jeroen Chua, Pedro F. Felzenszwalb
Abstract	We describe a general framework for probabilistic modeling of complex scenes and inference from ambiguous observations. The approach is motivated by applications in image analysis and is based on the use of priors defined by stochastic grammars. We define a class of grammars that capture relationships between the objects in a scene and provide important contextual cues for statistical inference. The distribution over scenes defined by a probabilistic scene grammar can be represented by a graphical model and this construction can be used for efficient inference with loopy belief propagation. We show experimental results with two different applications. One application involves the reconstruction of binary contour maps. Another application involves detecting and localizing faces in images. In both applications the same framework leads to robust inference algorithms that can effectively combine local information to reason about a scene.
Tasks
Published	2016-06-03
URL	https://arxiv.org/abs/1606.01307v3
PDF	https://arxiv.org/pdf/1606.01307v3.pdf
PWC	https://paperswithcode.com/paper/scene-grammars-factor-graphs-and-belief
Repo
Framework

A Multi-Pass Approach to Large-Scale Connectomics


Title	A Multi-Pass Approach to Large-Scale Connectomics
Authors	Yaron Meirovitch, Alexander Matveev, Hayk Saribekyan, David Budden, David Rolnick, Gergely Odor, Seymour Knowles-Barley, Thouis Raymond Jones, Hanspeter Pfister, Jeff William Lichtman, Nir Shavit
Abstract	The field of connectomics faces unprecedented “big data” challenges. To reconstruct neuronal connectivity, automated pixel-level segmentation is required for petabytes of streaming electron microscopy data. Existing algorithms provide relatively good accuracy but are unacceptably slow, and would require years to extract connectivity graphs from even a single cubic millimeter of neural tissue. Here we present a viable real-time solution, a multi-pass pipeline optimized for shared-memory multicore systems, capable of processing data at near the terabyte-per-hour pace of multi-beam electron microscopes. The pipeline makes an initial fast-pass over the data, and then makes a second slow-pass to iteratively correct errors in the output of the fast-pass. We demonstrate the accuracy of a sparse slow-pass reconstruction algorithm and suggest new methods for detecting morphological errors. Our fast-pass approach provided many algorithmic challenges, including the design and implementation of novel shallow convolutional neural nets and the parallelization of watershed and object-merging techniques. We use it to reconstruct, from image stack to skeletons, the full dataset of Kasthuri et al. (463 GB capturing 120,000 cubic microns) in a matter of hours on a single multicore machine rather than the weeks it has taken in the past on much larger distributed systems.
Tasks
Published	2016-12-07
URL	http://arxiv.org/abs/1612.02120v1
PDF	http://arxiv.org/pdf/1612.02120v1.pdf
PWC	https://paperswithcode.com/paper/a-multi-pass-approach-to-large-scale
Repo
Framework

Inverse Reinforcement Learning in Swarm Systems


Title	Inverse Reinforcement Learning in Swarm Systems
Authors	Adrian Šošić, Wasiur R. KhudaBukhsh, Abdelhak M. Zoubir, Heinz Koeppl
Abstract	Inverse reinforcement learning (IRL) has become a useful tool for learning behavioral models from demonstration data. However, IRL remains mostly unexplored for multi-agent systems. In this paper, we show how the principle of IRL can be extended to homogeneous large-scale problems, inspired by the collective swarming behavior of natural systems. In particular, we make the following contributions to the field: 1) We introduce the swarMDP framework, a sub-class of decentralized partially observable Markov decision processes endowed with a swarm characterization. 2) Exploiting the inherent homogeneity of this framework, we reduce the resulting multi-agent IRL problem to a single-agent one by proving that the agent-specific value functions in this model coincide. 3) To solve the corresponding control problem, we propose a novel heterogeneous learning scheme that is particularly tailored to the swarm setting. Results on two example systems demonstrate that our framework is able to produce meaningful local reward models from which we can replicate the observed global system dynamics.
Tasks
Published	2016-02-17
URL	http://arxiv.org/abs/1602.05450v2
PDF	http://arxiv.org/pdf/1602.05450v2.pdf
PWC	https://paperswithcode.com/paper/inverse-reinforcement-learning-in-swarm
Repo
Framework

Statistical Inference Using Mean Shift Denoising


Title	Statistical Inference Using Mean Shift Denoising
Authors	Yunhua Xiang, Yen-Chi Chen
Abstract	In this paper, we study how the mean shift algorithm can be used to denoise a dataset. We introduce a new framework to analyze the mean shift algorithm as a denoising approach by viewing the algorithm as an operator on a distribution function. We investigate how the mean shift algorithm changes the distribution and show that data points shifted by the mean shift concentrate around high density regions of the underlying density function. By using the mean shift as a denoising method, we enhance the performance of several clustering techniques, improve the power of two-sample tests, and obtain a new method for anomaly detection.
Tasks	Anomaly Detection, Denoising
Published	2016-10-13
URL	http://arxiv.org/abs/1610.03927v1
PDF	http://arxiv.org/pdf/1610.03927v1.pdf
PWC	https://paperswithcode.com/paper/statistical-inference-using-mean-shift
Repo
Framework

Emergent Predication Structure in Hidden State Vectors of Neural Readers


Title	Emergent Predication Structure in Hidden State Vectors of Neural Readers
Authors	Hai Wang, Takeshi Onishi, Kevin Gimpel, David McAllester
Abstract	A significant number of neural architectures for reading comprehension have recently been developed and evaluated on large cloze-style datasets. We present experiments supporting the emergence of “predication structure” in the hidden state vectors of these readers. More specifically, we provide evidence that the hidden state vectors represent atomic formulas $\Phi[c]$ where $\Phi$ is a semantic property (predicate) and $c$ is a constant symbol entity identifier.
Tasks	Reading Comprehension
Published	2016-11-23
URL	http://arxiv.org/abs/1611.07954v2
PDF	http://arxiv.org/pdf/1611.07954v2.pdf
PWC	https://paperswithcode.com/paper/emergent-predication-structure-in-hidden
Repo
Framework

Fuzzy Object-Oriented Dynamic Networks. II


Title	Fuzzy Object-Oriented Dynamic Networks. II
Authors	D. A. Terletskyi, A. I. Provotar
Abstract	This article generalizes object-oriented dynamic networks to the fuzzy case, which allows one to represent knowledge on objects and classes of objects that are fuzzy by nature and also to model their changes in time. Within the framework of the approach described, a mechanism is proposed that makes it possible to acquire new knowledge on the basis of basic knowledge and considerably differs from well-known methods used in existing models of knowledge representation. The approach is illustrated by an example of construction of a concrete fuzzy object-oriented dynamic network.
Tasks
Published	2016-02-04
URL	http://arxiv.org/abs/1602.01628v2
PDF	http://arxiv.org/pdf/1602.01628v2.pdf
PWC	https://paperswithcode.com/paper/fuzzy-object-oriented-dynamic-networks-ii
Repo
Framework

An Artificial Agent for Robust Image Registration


Title	An Artificial Agent for Robust Image Registration
Authors	Rui Liao, Shun Miao, Pierre de Tournemire, Sasa Grbic, Ali Kamen, Tommaso Mansi, Dorin Comaniciu
Abstract	3-D image registration, which involves aligning two or more images, is a critical step in a variety of medical applications from diagnosis to therapy. Image registration is commonly performed by optimizing an image matching metric as a cost function. However, this task is challenging due to the non-convex nature of the matching metric over the plausible registration parameter space and insufficient approaches for a robust optimization. As a result, current approaches are often customized to a specific problem and sensitive to image quality and artifacts. In this paper, we propose a completely different approach to image registration, inspired by how experts perform the task. We first cast the image registration problem as a “strategy learning” process, where the goal is to find the best sequence of motion actions (e.g. up, down, etc.) that yields image alignment. Within this approach, an artificial agent is learned, modeled using deep convolutional neural networks, with 3D raw image data as the input, and the next optimal action as the output. To cope with the dimensionality of the problem, we propose a greedy supervised approach for an end-to-end training, coupled with attention-driven hierarchical strategy. The resulting registration approach inherently encodes both a data-driven matching metric and an optimal registration strategy (policy). We demonstrate, on two 3-D/3-D medical image registration examples with drastically different nature of challenges, that the artificial agent outperforms several state-of-art registration methods by a large margin in terms of both accuracy and robustness.
Tasks	Image Registration, Medical Image Registration
Published	2016-11-30
URL	http://arxiv.org/abs/1611.10336v1
PDF	http://arxiv.org/pdf/1611.10336v1.pdf
PWC	https://paperswithcode.com/paper/an-artificial-agent-for-robust-image
Repo
Framework

Large age-gap face verification by feature injection in deep networks


Title	Large age-gap face verification by feature injection in deep networks
Authors	Simone Bianco
Abstract	This paper introduces a new method for face verification across large age gaps and also a dataset containing variations of age in the wild, the Large Age-Gap (LAG) dataset, with images ranging from child/young to adult/old. The proposed method exploits a deep convolutional neural network (DCNN) pre-trained for the face recognition task on a large dataset and then fine-tuned for the large age-gap face verification task. Finetuning is performed in a Siamese architecture using a contrastive loss function. A feature injection layer is introduced to boost verification accuracy, showing the ability of the DCNN to learn a similarity metric leveraging external features. Experimental results on the LAG dataset show that our method is able to outperform the face verification solutions in the state of the art considered.
Tasks	Age-Invariant Face Recognition, Face Recognition, Face Verification
Published	2016-02-19
URL	http://arxiv.org/abs/1602.06149v1
PDF	http://arxiv.org/pdf/1602.06149v1.pdf
PWC	https://paperswithcode.com/paper/large-age-gap-face-verification-by-feature
Repo
Framework

Action Recognition with Joint Attention on Multi-Level Deep Features


Title	Action Recognition with Joint Attention on Multi-Level Deep Features
Authors	Jialin Wu, Gu Wang, Wukui Yang, Xiangyang Ji
Abstract	We propose a novel deep supervised neural network for the task of action recognition in videos, which implicitly takes advantage of visual tracking and shares the robustness of both deep Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN). In our method, a multi-branch model is proposed to suppress noise from background jitters. Specifically, we firstly extract multi-level deep features from deep CNNs and feed them into 3d-convolutional network. After that we feed those feature cubes into our novel joint LSTM module to predict labels and to generate attention regularization. We evaluate our model on two challenging datasets: UCF101 and HMDB51. The results show that our model achieves the state-of-art by only using convolutional features.
Tasks	Action Recognition In Videos, Temporal Action Localization, Visual Tracking
Published	2016-07-09
URL	http://arxiv.org/abs/1607.02556v1
PDF	http://arxiv.org/pdf/1607.02556v1.pdf
PWC	https://paperswithcode.com/paper/action-recognition-with-joint-attention-on
Repo
Framework