April 2, 2020

3118 words 15 mins read

Paper Group ANR 319

Feature-map-level Online Adversarial Knowledge Distillation. Learning a Probabilistic Strategy for Computational Imaging Sensor Selection. StarAI: Reducing incompleteness in the game of Bridge using PLP. Multimodal Data Fusion based on the Global Workspace Theory. Curved Buildings Reconstruction from Airborne LiDAR Data by Matching and Deforming Ge …

Feature-map-level Online Adversarial Knowledge Distillation


Title	Feature-map-level Online Adversarial Knowledge Distillation
Authors	Inseop Chung, SeongUk Park, Jangho Kim, Nojun Kwak
Abstract	Feature maps contain rich information about image intensity and spatial correlation. However, previous online knowledge distillation methods only utilize the class probabilities. Thus in this paper, we propose an online knowledge distillation method that transfers not only the knowledge of the class probabilities but also that of the feature map using the adversarial training framework. We train multiple networks simultaneously by employing discriminators to distinguish the feature map distributions of different networks. Each network has its corresponding discriminator which discriminates the feature map from its own as fake while classifying that of the other network as real. By training a network to fool the corresponding discriminator, it can learn the other network’s feature map distribution. We show that our method performs better than the conventional direct alignment method such as L1 and is more suitable for online distillation. Also, we propose a novel cyclic learning scheme for training more than two networks together. We have applied our method to various network architectures on the classification task and discovered a significant improvement of performance especially in the case of training a pair of a small network and a large one.
Tasks
Published	2020-02-05
URL	https://arxiv.org/abs/2002.01775v1
PDF	https://arxiv.org/pdf/2002.01775v1.pdf
PWC	https://paperswithcode.com/paper/feature-map-level-online-adversarial-1
Repo
Framework

Learning a Probabilistic Strategy for Computational Imaging Sensor Selection


Title	Learning a Probabilistic Strategy for Computational Imaging Sensor Selection
Authors	He Sun, Adrian V. Dalca, Katherine L. Bouman
Abstract	Optimized sensing is important for computational imaging in low-resource environments, when images must be recovered from severely limited measurements. In this paper, we propose a physics-constrained, fully differentiable, autoencoder that learns a probabilistic sensor-sampling strategy for optimized sensor design. The proposed method learns a system’s preferred sampling distribution that characterizes the correlations between different sensor selections as a binary, fully-connected Ising model. The learned probabilistic model is achieved by using a Gibbs sampling inspired network architecture, and is trained end-to-end with a reconstruction network for efficient co-design. The proposed framework is applicable to sensor selection problems in a variety of computational imaging applications. In this paper, we demonstrate the approach in the context of a very-long-baseline-interferometry (VLBI) array design task, where sensor correlations and atmospheric noise present unique challenges. We demonstrate results broadly consistent with expectation, and draw attention to particular structures preferred in the telescope array geometry that can be leveraged to plan future observations and design array expansions.
Tasks
Published	2020-03-23
URL	https://arxiv.org/abs/2003.10424v1
PDF	https://arxiv.org/pdf/2003.10424v1.pdf
PWC	https://paperswithcode.com/paper/learning-a-probabilistic-strategy-for
Repo
Framework

StarAI: Reducing incompleteness in the game of Bridge using PLP


Title	StarAI: Reducing incompleteness in the game of Bridge using PLP
Authors	J Li, S Thepaut, V Ventos
Abstract	Bridge is a trick-taking card game requiring the ability to evaluate probabilities since it is a game of incomplete information where each player only sees its cards. In order to choose a strategy, a player needs to gather information about the hidden cards in the other players’ hand. We present a methodology allowing us to model a part of card playing in Bridge using Probabilistic Logic Programming.
Tasks
Published	2020-01-22
URL	https://arxiv.org/abs/2001.08193v1
PDF	https://arxiv.org/pdf/2001.08193v1.pdf
PWC	https://paperswithcode.com/paper/starai-reducing-incompleteness-in-the-game-of
Repo
Framework

Multimodal Data Fusion based on the Global Workspace Theory


Title	Multimodal Data Fusion based on the Global Workspace Theory
Authors	Cong Bao, Zafeirios Fountas, Temitayo Olugbade, Nadia Bianchi-Berthouze
Abstract	We propose a novel neural network architecture, named the Global Workspace Network (GWN), that addresses the challenge of dynamic uncertainties in multimodal data fusion. The GWN is inspired by the well-established Global Workspace Theory from cognitive science. We implement it as a model of attention, between multiple modalities, that evolves through time. The GWN achieved F1 score of 0.92, averaged over two classes, for the discrimination between patient and healthy participants, based on the multimodal EmoPain dataset captured from people with chronic pain and healthy people performing different types of exercise movements in unconstrained settings. In this task, the GWN significantly outperformed a vanilla architecture. It additionally outperformed the vanilla model in further classification of three pain levels for a patient (average F1 score = 0.75) based on the EmoPain dataset. We further provide extensive analysis of the behaviour of GWN and its ability to deal with uncertainty in multimodal data.
Tasks
Published	2020-01-26
URL	https://arxiv.org/abs/2001.09485v1
PDF	https://arxiv.org/pdf/2001.09485v1.pdf
PWC	https://paperswithcode.com/paper/multimodal-data-fusion-based-on-the-global
Repo
Framework

Curved Buildings Reconstruction from Airborne LiDAR Data by Matching and Deforming Geometric Primitives


Title	Curved Buildings Reconstruction from Airborne LiDAR Data by Matching and Deforming Geometric Primitives
Authors	Jingwei Song, Shaobo Xia, Jun Wang, Dong Chen
Abstract	Airborne LiDAR (Light Detection and Ranging) data is widely applied in building reconstruction, with studies reporting success in typical buildings. However, the reconstruction of curved buildings remains an open research problem. To this end, we propose a new framework for curved building reconstruction via assembling and deforming geometric primitives. The input LiDAR point cloud are first converted into contours where individual buildings are identified. After recognizing geometric units (primitives) from building contours, we get initial models by matching basic geometric primitives to these primitives. To polish assembly models, we employ a warping field for model refinements. Specifically, an embedded deformation (ED) graph is constructed via downsampling the initial model. Then, the point-to-model displacements are minimized by adjusting node parameters in the ED graph based on our objective function. The presented framework is validated on several highly curved buildings collected by various LiDAR in different cities. The experimental results, as well as accuracy comparison, demonstrate the advantage and effectiveness of our method. {The new insight attributes to an efficient reconstruction manner.} Moreover, we prove that the primitive-based framework significantly reduces the data storage to 10-20 percent of classical mesh models.
Tasks
Published	2020-03-22
URL	https://arxiv.org/abs/2003.09934v1
PDF	https://arxiv.org/pdf/2003.09934v1.pdf
PWC	https://paperswithcode.com/paper/curved-buildings-reconstruction-from-airborne
Repo
Framework

Evaluation of Cross-View Matching to Improve Ground Vehicle Localization with Aerial Perception


Title	Evaluation of Cross-View Matching to Improve Ground Vehicle Localization with Aerial Perception
Authors	Deeksha Dixit, Pratap Tokekar
Abstract	Cross-view matching refers to the problem of finding the closest match to a given query ground-view image to one from a database of aerial images. If the aerial images are geotagged, then the closest matching aerial image can be used to localize the query ground-view image. Recently, due to the success of deep learning methods, a number of cross-view matching techniques have been proposed. These techniques perform well for the matching of isolated query images. In this paper, we evaluate cross-view matching for the task of localizing a ground vehicle over a longer trajectory. We use the cross-view matching module as a sensor measurement fused with a particle filter. We evaluate the performance of this method using a city-wide dataset collected in photorealistic simulation using five parameters: height of aerial images, the pitch of the aerial camera mount, field-of-view of ground camera, measurement model and resampling strategy for the particles in the particle filter.
Tasks
Published	2020-03-13
URL	https://arxiv.org/abs/2003.06515v3
PDF	https://arxiv.org/pdf/2003.06515v3.pdf
PWC	https://paperswithcode.com/paper/evaluation-of-cross-view-matching-to-improve
Repo
Framework

Evaluating Low-Resource Machine Translation between Chinese and Vietnamese with Back-Translation


Title	Evaluating Low-Resource Machine Translation between Chinese and Vietnamese with Back-Translation
Authors	Hongzheng Li, Heyan Huang
Abstract	Back translation (BT) has been widely used and become one of standard techniques for data augmentation in Neural Machine Translation (NMT), BT has proven to be helpful for improving the performance of translation effectively, especially for low-resource scenarios. While most works related to BT mainly focus on European languages, few of them study languages in other areas around the world. In this paper, we investigate the impacts of BT on Asia language translations between the extremely low-resource Chinese and Vietnamese language pair. We evaluate and compare the effects of different sizes of synthetic data on both NMT and Statistical Machine Translation (SMT) models for Chinese to Vietnamese and Vietnamese to Chinese, with character-based and word-based settings. Some conclusions from previous works are partially confirmed and we also draw some other interesting findings and conclusions, which are beneficial to understand BT further.
Tasks	Data Augmentation, Machine Translation
Published	2020-03-04
URL	https://arxiv.org/abs/2003.02197v2
PDF	https://arxiv.org/pdf/2003.02197v2.pdf
PWC	https://paperswithcode.com/paper/evaluating-low-resource-machine-translation
Repo
Framework

Solving high-dimensional eigenvalue problems using deep neural networks: A diffusion Monte Carlo like approach


Title	Solving high-dimensional eigenvalue problems using deep neural networks: A diffusion Monte Carlo like approach
Authors	Jiequn Han, Jianfeng Lu, Mo Zhou
Abstract	We propose a new method to solve eigenvalue problems for linear and semilinear second order differential operators in high dimensions based on deep neural networks. The eigenvalue problem is reformulated as a fixed point problem of the semigroup flow induced by the operator, whose solution can be represented by Feynman-Kac formula in terms of forward-backward stochastic differential equations. The method shares a similar spirit with diffusion Monte Carlo but augments a direct approximation to the eigenfunction through neural-network ansatz. The criterion of fixed point provides a natural loss function to search for parameters via optimization. Our approach is able to provide accurate eigenvalue and eigenfunction approximations in several numerical examples, including Fokker-Planck operator, linear and nonlinear Schr"odinger operators in high dimensions.
Tasks
Published	2020-02-07
URL	https://arxiv.org/abs/2002.02600v1
PDF	https://arxiv.org/pdf/2002.02600v1.pdf
PWC	https://paperswithcode.com/paper/solving-high-dimensional-eigenvalue-problems
Repo
Framework

Perception as prediction using general value functions in autonomous driving applications


Title	Perception as prediction using general value functions in autonomous driving applications
Authors	Daniel Graves, Kasra Rezaee, Sean Scheideman
Abstract	We propose and demonstrate a framework called perception as prediction for autonomous driving that uses general value functions (GVFs) to learn predictions. Perception as prediction learns data-driven predictions relating to the impact of actions on the agent’s perception of the world. It also provides a data-driven approach to predict the impact of the anticipated behavior of other agents on the world without explicitly learning their policy or intentions. We demonstrate perception as prediction by learning to predict an agent’s front safety and rear safety with GVFs, which encapsulate anticipation of the behavior of the vehicle in front and in the rear, respectively. The safety predictions are learned through random interactions in a simulated environment containing other agents. We show that these predictions can be used to produce similar control behavior to an LQR-based controller in an adaptive cruise control problem as well as provide advanced warning when the vehicle behind is approaching dangerously. The predictions are compact policy-based predictions that support prediction of the long term impact on safety when following a given policy. We analyze two controllers that use the learned predictions in a racing simulator to understand the value of the predictions and demonstrate their use in the real-world on a Clearpath Jackal robot and an autonomous vehicle platform.
Tasks	Autonomous Driving
Published	2020-01-24
URL	https://arxiv.org/abs/2001.09113v1
PDF	https://arxiv.org/pdf/2001.09113v1.pdf
PWC	https://paperswithcode.com/paper/perception-as-prediction-using-general-value
Repo
Framework

Online Learning with Imperfect Hints


Title	Online Learning with Imperfect Hints
Authors	Aditya Bhaskara, Ashok Cutkosky, Ravi Kumar, Manish Purohit
Abstract	We consider a variant of the classical online linear optimization problem in which at every step, the online player receives a “hint” vector before choosing the action for that round. Rather surprisingly, it was shown that if the hint vector is guaranteed to have a positive correlation with the cost vector, then the online player can achieve a regret of $O(\log T)$, thus significantly improving over the $O(\sqrt{T})$ regret in the general setting. However, the result and analysis require the correlation property at \emph{all} time steps, thus raising the natural question: can we design online learning algorithms that are resilient to bad hints? In this paper we develop algorithms and nearly matching lower bounds for online learning with imperfect directional hints. Our algorithms are oblivious to the quality of the hints, and the regret bounds interpolate between the always-correlated hints case and the no-hints case. Our results also generalize, simplify, and improve upon previous results on optimistic regret bounds, which can be viewed as an additive version of hints.
Tasks
Published	2020-02-11
URL	https://arxiv.org/abs/2002.04726v1
PDF	https://arxiv.org/pdf/2002.04726v1.pdf
PWC	https://paperswithcode.com/paper/online-learning-with-imperfect-hints
Repo
Framework

Hybrid Generative-Retrieval Transformers for Dialogue Domain Adaptation


Title	Hybrid Generative-Retrieval Transformers for Dialogue Domain Adaptation
Authors	Igor Shalyminov, Alessandro Sordoni, Adam Atkinson, Hannes Schulz
Abstract	Domain adaptation has recently become a key problem in dialogue systems research. Deep learning, while being the preferred technique for modeling such systems, works best given massive training data. However, in the real-world scenario, such resources aren’t available for every new domain, so the ability to train with a few dialogue examples can be considered essential. Pre-training on large data sources and adapting to the target data has become the standard method for few-shot problems within the deep learning framework. In this paper, we present the winning entry at the fast domain adaptation task of DSTC8, a hybrid generative-retrieval model based on GPT-2 fine-tuned to the multi-domain MetaLWOz dataset. Robust and diverse in response generation, our model uses retrieval logic as a fallback, being SoTA on MetaLWOz in human evaluation (>4% improvement over the 2nd place system) and attaining competitive generalization performance in adaptation to the unseen MultiWOZ dataset.
Tasks	Domain Adaptation
Published	2020-03-03
URL	https://arxiv.org/abs/2003.01680v2
PDF	https://arxiv.org/pdf/2003.01680v2.pdf
PWC	https://paperswithcode.com/paper/hybrid-generative-retrieval-transformers-for
Repo
Framework

Improving Performance in Reinforcement Learning by Breaking Generalization in Neural Networks


Title	Improving Performance in Reinforcement Learning by Breaking Generalization in Neural Networks
Authors	Sina Ghiassian, Banafsheh Rafiee, Yat Long Lo, Adam White
Abstract	Reinforcement learning systems require good representations to work well. For decades practical success in reinforcement learning was limited to small domains. Deep reinforcement learning systems, on the other hand, are scalable, not dependent on domain specific prior knowledge and have been successfully used to play Atari, in 3D navigation from pixels, and to control high degree of freedom robots. Unfortunately, the performance of deep reinforcement learning systems is sensitive to hyper-parameter settings and architecture choices. Even well tuned systems exhibit significant instability both within a trial and across experiment replications. In practice, significant expertise and trial and error are usually required to achieve good performance. One potential source of the problem is known as catastrophic interference: when later training decreases performance by overriding previous learning. Interestingly, the powerful generalization that makes Neural Networks (NN) so effective in batch supervised learning might explain the challenges when applying them in reinforcement learning tasks. In this paper, we explore how online NN training and interference interact in reinforcement learning. We find that simply re-mapping the input observations to a high-dimensional space improves learning speed and parameter sensitivity. We also show this preprocessing reduces interference in prediction tasks. More practically, we provide a simple approach to NN training that is easy to implement, and requires little additional computation. We demonstrate that our approach improves performance in both prediction and control with an extensive batch of experiments in classic control domains.
Tasks
Published	2020-03-16
URL	https://arxiv.org/abs/2003.07417v1
PDF	https://arxiv.org/pdf/2003.07417v1.pdf
PWC	https://paperswithcode.com/paper/improving-performance-in-reinforcement
Repo
Framework

Semi-supervised learning method based on predefined evenly-distributed class centroids


Title	Semi-supervised learning method based on predefined evenly-distributed class centroids
Authors	Qiuyu Zhu, Tiantian Li
Abstract	Compared to supervised learning, semi-supervised learning reduces the dependence of deep learning on a large number of labeled samples. In this work, we use a small number of labeled samples and perform data augmentation on unlabeled samples to achieve image classification. Our method constrains all samples to the predefined evenly-distributed class centroids (PEDCC) by the corresponding loss function. Specifically, the PEDCC-Loss for labeled samples, and the maximum mean discrepancy loss for unlabeled samples are used to make the feature distribution closer to the distribution of PEDCC. Our method ensures that the inter-class distance is large and the intra-class distance is small enough to make the classification boundaries between different classes clearer. Meanwhile, for unlabeled samples, we also use KL divergence to constrain the consistency of the network predictions between unlabeled and augmented samples. Our semi-supervised learning method achieves the state-of-the-art results, with 4000 labeled samples on CIFAR10 and 1000 labeled samples on SVHN, and the accuracy is 95.10% and 97.58% respectively.
Tasks	Data Augmentation, Image Classification
Published	2020-01-13
URL	https://arxiv.org/abs/2001.04092v1
PDF	https://arxiv.org/pdf/2001.04092v1.pdf
PWC	https://paperswithcode.com/paper/semi-supervised-learning-method-based-on
Repo
Framework

Face Phylogeny Tree Using Basis Functions


Title	Face Phylogeny Tree Using Basis Functions
Authors	Sudipta Banerjee, Arun Ross
Abstract	Photometric transformations, such as brightness and contrast adjustment, can be applied to a face image repeatedly creating a set of near-duplicate images. Identifying the original image from a set of such near-duplicates and deducing the relationship between them are important in the context of digital image forensics. This is commonly done by generating an image phylogeny tree \textemdash \hspace{0.08cm} a hierarchical structure depicting the relationship between a set of near-duplicate images. In this work, we utilize three different families of basis functions to model pairwise relationships between near-duplicate images. The basis functions used in this work are orthogonal polynomials, wavelet basis functions and radial basis functions. We perform extensive experiments to assess the performance of the proposed method across three different modalities, namely, face, fingerprint and iris images; across different image phylogeny tree configurations; and across different types of photometric transformations. We also utilize the same basis functions to model geometric transformations and deep-learning based transformations. We also perform extensive analysis of each basis function with respect to its ability to model arbitrary transformations and to distinguish between the original and the transformed images. Finally, we utilize the concept of approximate von Neumann graph entropy to explain the success and failure cases of the proposed IPT generation algorithm. Experiments indicate that the proposed algorithm generalizes well across different scenarios thereby suggesting the merits of using basis functions to model the relationship between photometrically and geometrically modified images.
Tasks
Published	2020-02-21
URL	https://arxiv.org/abs/2002.09068v2
PDF	https://arxiv.org/pdf/2002.09068v2.pdf
PWC	https://paperswithcode.com/paper/face-phylogeny-tree-using-basis-functions
Repo
Framework

Durocmien: A deep framework for duroc skeleton extraction in constraint environment


Title	Durocmien: A deep framework for duroc skeleton extraction in constraint environment
Authors	Akif Quddus Khan, Salman Khan
Abstract	Farm animal behavior analysis is a crucial tasks for the industrial farming. In an indoor farm setting, extracting Key joints of animal is essential for tracking the animal for longer period of time. In this paper, we proposed a deep network named DUROCMIEN that exploit transfer learning to trained the network for the Duroc, a domestic breed of pig, an end to end fashion. The backbone of the architecture is based on hourglass stacked dense-net. In order to train the network, key frames are selected from the test data using K-mean sampler. In total, 9 Keypoints are annotated that gives a brief detailed behavior analysis in the farm setting. Extensive experiments are conducted and the quantitative results show that the network has the potential of increasing the tracking performance by a substantial margin.
Tasks	Transfer Learning
Published	2020-01-17
URL	https://arxiv.org/abs/2002.03727v1
PDF	https://arxiv.org/pdf/2002.03727v1.pdf
PWC	https://paperswithcode.com/paper/durocmien-a-deep-framework-for-duroc-skeleton
Repo
Framework