January 28, 2020

2858 words 14 mins read

Paper Group ANR 1033

Sample-efficient Adversarial Imitation Learning from Observation. RIDM: Reinforced Inverse Dynamics Modeling for Learning from a Single Observed Demonstration. Neural Architecture Search on Acoustic Scene Classification. Predicting Tomorrow’s Headline using Today’s Twitter Deliberations. Prediction of wall-bounded turbulence from wall quantities us …

Sample-efficient Adversarial Imitation Learning from Observation


Title	Sample-efficient Adversarial Imitation Learning from Observation
Authors	Faraz Torabi, Sean Geiger, Garrett Warnell, Peter Stone
Abstract	Imitation from observation is the framework of learning tasks by observing demonstrated state-only trajectories. Recently, adversarial approaches have achieved significant performance improvements over other methods for imitating complex behaviors. However, these adversarial imitation algorithms often require many demonstration examples and learning iterations to produce a policy that is successful at imitating a demonstrator’s behavior. This high sample complexity often prohibits these algorithms from being deployed on physical robots. In this paper, we propose an algorithm that addresses the sample inefficiency problem by utilizing ideas from trajectory centric reinforcement learning algorithms. We test our algorithm and conduct experiments using an imitation task on a physical robot arm and its simulated version in Gazebo and will show the improvement in learning rate and efficiency.
Tasks	Imitation Learning
Published	2019-06-18
URL	https://arxiv.org/abs/1906.07374v1
PDF	https://arxiv.org/pdf/1906.07374v1.pdf
PWC	https://paperswithcode.com/paper/sample-efficient-adversarial-imitation
Repo
Framework

RIDM: Reinforced Inverse Dynamics Modeling for Learning from a Single Observed Demonstration


Title	RIDM: Reinforced Inverse Dynamics Modeling for Learning from a Single Observed Demonstration
Authors	Brahma S. Pavse, Faraz Torabi, Josiah P. Hanna, Garrett Warnell, Peter Stone
Abstract	Imitation learning has long been an approach to alleviate the tractability issues that arise in reinforcement learning. However, most literature makes several assumptions such as access to the expert’s actions, availability of many expert demonstrations, and injection of task-specific domain knowledge into the learning process. We propose reinforced inverse dynamics modeling (RIDM), a method of combining reinforcement learning and imitation from observation (IfO) to perform imitation using a single expert demonstration, with no access to the expert’s actions, and with little task-specific domain knowledge. Given only a single set of the expert’s raw states, such as joint angles in a robot control task, at each time-step, we learn an inverse dynamics model to produce the necessary low-level actions, such as torques, to transition from one state to the next such that the reward from the environment is maximized. We demonstrate that RIDM outperforms other techniques when we apply the same constraints on the other methods on six domains of the MuJoCo simulator and for two different robot soccer tasks for two experts from the RoboCup 3D simulation league on the SimSpark simulator.
Tasks	Imitation Learning
Published	2019-06-18
URL	https://arxiv.org/abs/1906.07372v3
PDF	https://arxiv.org/pdf/1906.07372v3.pdf
PWC	https://paperswithcode.com/paper/ridm-reinforced-inverse-dynamics-modeling-for
Repo
Framework

Neural Architecture Search on Acoustic Scene Classification


Title	Neural Architecture Search on Acoustic Scene Classification
Authors	Jixiang Li, Chuming Liang, Bo Zhang, Zhao Wang, Fei Xiang, Xiangxiang Chu
Abstract	Convolutional neural networks are widely adopted in Acoustic Scene Classification (ASC) tasks, but they generally carry a heavy computational burden. In this work, we propose a lightweight yet high-performing baseline network inspired by MobileNetV2, which replaces square convolutional kernels with unidirectional ones to extract features alternately in temporal and frequency dimensions. Furthermore, we explore a dynamic architecture space built on the basis of the proposed baseline with the recent Neural Architecture Search (NAS) paradigm, which first trains a supernet that incorporates all candidate networks and then applies a well-known evolutionary algorithm NSGA-II to discover more efficient networks with higher accuracy and lower computational cost. Experimental results demonstrate that our searched network is competent in ASC tasks, which achieves 90.3% F1-score on the DCASE2018 task 5 evaluation set, marking a new state-of-the-art performance while saving 25% of FLOPs compared to our baseline network.
Tasks	Acoustic Scene Classification, Neural Architecture Search, Scene Classification
Published	2019-12-30
URL	https://arxiv.org/abs/1912.12825v1
PDF	https://arxiv.org/pdf/1912.12825v1.pdf
PWC	https://paperswithcode.com/paper/neural-architecture-search-on-acoustic-scene
Repo
Framework

Predicting Tomorrow’s Headline using Today’s Twitter Deliberations


Title	Predicting Tomorrow’s Headline using Today’s Twitter Deliberations
Authors	Roshni Chakraborty, Abhijeet Kharat, Apalak Khatua, Sourav Kumar Dandapat, Joydeep Chandra
Abstract	Predicting the popularity of news article is a challenging task. Existing literature mostly focused on article contents and polarity to predict popularity. However, existing research has not considered the users’ preference towards a particular article. Understanding users’ preference is an important aspect for predicting the popularity of news articles. Hence, we consider the social media data, from the Twitter platform, to address this research gap. In our proposed model, we have considered the users’ involvement as well as the users’ reaction towards an article to predict the popularity of the article. In short, we are predicting tomorrow’s headline by probing today’s Twitter discussion. We have considered 300 political news article from the New York Post, and our proposed approach has outperformed other baseline models.
Tasks
Published	2019-01-27
URL	http://arxiv.org/abs/1901.09334v1
PDF	http://arxiv.org/pdf/1901.09334v1.pdf
PWC	https://paperswithcode.com/paper/predicting-tomorrows-headline-using-todays
Repo
Framework

Prediction of wall-bounded turbulence from wall quantities using convolutional neural networks


Title	Prediction of wall-bounded turbulence from wall quantities using convolutional neural networks
Authors	L. Guastoni, M. P. Encinar, P. Schlatter, H. Azizpour, R. Vinuesa
Abstract	A fully-convolutional neural-network model is used to predict the streamwise velocity fields at several wall-normal locations by taking as input the streamwise and spanwise wall-shear-stress planes in a turbulent open channel flow. The training data are generated by performing a direct numerical simulation (DNS) at a friction Reynolds number of $Re_{\tau}=180$. Various networks are trained for predictions at three inner-scaled locations ($y^+ = 15,~30,~50$) and for different time steps between input samples $\Delta t^{+}_{s}$. The inherent non-linearity of the neural-network model enables a better prediction capability than linear methods, with a lower error in both the instantaneous flow fields and turbulent statistics. Using a dataset with higher $\Delta t^+_{s}$ improves the generalization at all the considered wall-normal locations, as long as the network capacity is sufficient to generalize over the dataset. The use of a multiple-output network, with parallel dedicated branches for two wall-normal locations, does not provide any improvement over two separated single-output networks, other than a moderate saving in training time. Training time can be effectively reduced, by a factor of 4, via a transfer learning method that initializes the network parameters using the optimized parameters of a previously-trained network.
Tasks	Transfer Learning
Published	2019-12-30
URL	https://arxiv.org/abs/1912.12969v1
PDF	https://arxiv.org/pdf/1912.12969v1.pdf
PWC	https://paperswithcode.com/paper/prediction-of-wall-bounded-turbulence-from
Repo
Framework

Totally Deep Support Vector Machines


Title	Totally Deep Support Vector Machines
Authors	Hichem Sahbi
Abstract	Support vector machines (SVMs) have been successful in solving many computer vision tasks including image and video category recognition especially for small and mid-scale training problems. The principle of these non-parametric models is to learn hyperplanes that separate data belonging to different classes while maximizing their margins. However, SVMs constrain the learned hyperplanes to lie in the span of support vectors, fixed/taken from training data, and this reduces their representational power and may lead to limited generalization performances. In this paper, we relax this constraint and allow the support vectors to be learned (instead of being fixed/taken from training data) in order to better fit a given classification task. Our approach, referred to as deep total variation support vector machines, is parametric and relies on a novel deep architecture that learns not only the SVM and the kernel parameters but also the support vectors, resulting into highly effective classifiers. We also show (under a particular setting of the activation functions in this deep architecture) that a large class of kernels and their combinations can be learned. Experiments conducted on the challenging task of skeleton-based action recognition show the outperformance of our deep total variation SVMs w.r.t different baselines as well as the related work.
Tasks	Skeleton Based Action Recognition
Published	2019-12-12
URL	https://arxiv.org/abs/1912.05864v1
PDF	https://arxiv.org/pdf/1912.05864v1.pdf
PWC	https://paperswithcode.com/paper/totally-deep-support-vector-machines
Repo
Framework

Asynchronous Averaging of Gait Cycles for Classification of Gait and Device Modes


Title	Asynchronous Averaging of Gait Cycles for Classification of Gait and Device Modes
Authors	Parinaz Kasebzadeh, Gustaf Hendeby, Fredrik Gustafsson
Abstract	An approach for computing unique gait signature using measurements collected from body-worn inertial measurement units (IMUs) is proposed. The gait signature represents one full cycle of the human gait, and is suitable for off-line or on-line classification of the gait mode. The signature can also be used to jointly classify the gait mode and the device mode. The device mode identifies how the IMU-equipped device is being carried by the user. The method is based on precise segmentation and resampling of the measured IMU signal, as an initial step, further tuned by minimizing the variability of the obtained signature within each gait cycle. Finally, a Fourier series expansion of the gait signature is introduced which provides a low-dimensional feature vector well suited for classification purposes. The proposed method is evaluated on a large dataset involving several subjects, each one containing two different gait modes and four different device modes. The gait signatures enable a high classification rate for each step cycle.
Tasks
Published	2019-07-04
URL	https://arxiv.org/abs/1907.02329v2
PDF	https://arxiv.org/pdf/1907.02329v2.pdf
PWC	https://paperswithcode.com/paper/asynchronous-averaging-of-gait-cycles-for
Repo
Framework

Goal-conditioned Imitation Learning


Title	Goal-conditioned Imitation Learning
Authors	Yiming Ding, Carlos Florensa, Mariano Phielipp, Pieter Abbeel
Abstract	Designing rewards for Reinforcement Learning (RL) is challenging because it needs to convey the desired task, be efficient to optimize, and be easy to compute. The latter is particularly problematic when applying RL to robotics, where detecting whether the desired configuration is reached might require considerable supervision and instrumentation. Furthermore, we are often interested in being able to reach a wide range of configurations, hence setting up a different reward every time might be unpractical. Methods like Hindsight Experience Replay (HER) have recently shown promise to learn policies able to reach many goals, without the need of a reward. Unfortunately, without tricks like resetting to points along the trajectory, HER might require many samples to discover how to reach certain areas of the state-space. In this work we investigate different approaches to incorporate demonstrations to drastically speed up the convergence to a policy able to reach any goal, also surpassing the performance of an agent trained with other Imitation Learning algorithms. Furthermore, we show our method can also be used when the available expert trajectories do not contain the actions, which can leverage kinesthetic or third person demonstration. The code is available at https://sites.google.com/view/goalconditioned-il/.
Tasks	Imitation Learning
Published	2019-06-13
URL	https://arxiv.org/abs/1906.05838v2
PDF	https://arxiv.org/pdf/1906.05838v2.pdf
PWC	https://paperswithcode.com/paper/goal-conditioned-imitation-learning
Repo
Framework

Harnessing the Vulnerability of Latent Layers in Adversarially Trained Models


Title	Harnessing the Vulnerability of Latent Layers in Adversarially Trained Models
Authors	Mayank Singh, Abhishek Sinha, Nupur Kumari, Harshitha Machiraju, Balaji Krishnamurthy, Vineeth N Balasubramanian
Abstract	Neural networks are vulnerable to adversarial attacks – small visually imperceptible crafted noise which when added to the input drastically changes the output. The most effective method of defending against these adversarial attacks is to use the methodology of adversarial training. We analyze the adversarially trained robust models to study their vulnerability against adversarial attacks at the level of the latent layers. Our analysis reveals that contrary to the input layer which is robust to adversarial attack, the latent layer of these robust models are highly susceptible to adversarial perturbations of small magnitude. Leveraging this information, we introduce a new technique Latent Adversarial Training (LAT) which comprises of fine-tuning the adversarially trained models to ensure the robustness at the feature layers. We also propose Latent Attack (LA), a novel algorithm for construction of adversarial examples. LAT results in minor improvement in test accuracy and leads to a state-of-the-art adversarial accuracy against the universal first-order adversarial PGD attack which is shown for the MNIST, CIFAR-10, CIFAR-100 datasets.
Tasks	Adversarial Attack
Published	2019-05-13
URL	https://arxiv.org/abs/1905.05186v2
PDF	https://arxiv.org/pdf/1905.05186v2.pdf
PWC	https://paperswithcode.com/paper/harnessing-the-vulnerability-of-latent-layers
Repo
Framework

Deep segmentation networks predict survival of non-small cell lung cancer


Title	Deep segmentation networks predict survival of non-small cell lung cancer
Authors	Stephen Baek, Yusen He, Bryan G. Allen, John M. Buatti, Brian J. Smith, Ling Tong, Zhiyu Sun, Jia Wu, Maximilian Diehn, Billy W. Loo, Kristin A. Plichta, Steven N. Seyedin, Maggie Gannon, Katherine R. Cabel, Yusung Kim, Xiaodong Wu
Abstract	Non-small-cell lung cancer (NSCLC) represents approximately 80-85% of lung cancer diagnoses and is the leading cause of cancer-related death worldwide. Recent studies indicate that image-based radiomics features from positron emission tomography-computed tomography (PET/CT) images have predictive power on NSCLC outcomes. To this end, easily calculated functional features such as the maximum and the mean of standard uptake value (SUV) and total lesion glycolysis (TLG) are most commonly used for NSCLC prognostication, but their prognostic value remains controversial. Meanwhile, convolutional neural networks (CNN) are rapidly emerging as a new premise for cancer image analysis, with significantly enhanced predictive power compared to other hand-crafted radiomics features. Here we show that CNN trained to perform the tumor segmentation task, with no other information than physician contours, identify a rich set of survival-related image features with remarkable prognostic value. In a retrospective study on 96 NSCLC patients before stereotactic-body radiotherapy (SBRT), we found that the CNN segmentation algorithm (U-Net) trained for tumor segmentation in PET/CT images, contained features having strong correlation with 2- and 5-year overall and disease-specific survivals. The U-net algorithm has not seen any other clinical information (e.g. survival, age, smoking history) than the images and the corresponding tumor contours provided by physicians. Furthermore, through visualization of the U-Net, we also found convincing evidence that the regions of progression appear to match with the regions where the U-Net features identified patterns that predicted higher likelihood of death. We anticipate our findings will be a starting point for more sophisticated non-intrusive patient specific cancer prognosis determination.
Tasks
Published	2019-03-26
URL	https://arxiv.org/abs/1903.11593v2
PDF	https://arxiv.org/pdf/1903.11593v2.pdf
PWC	https://paperswithcode.com/paper/what-does-ai-see-deep-segmentation-networks
Repo
Framework

Bitopological Duality for Algebras of Fittings logic and Natural Duality extension


Title	Bitopological Duality for Algebras of Fittings logic and Natural Duality extension
Authors	Litan Kumar Das, Kumar Sankar Ray
Abstract	In this paper, we investigate a bitopological duality for algebras of Fitting’s multi-valued logic. We also extend the natural duality theory for $\mathbb{ISP_I}(\mathcal{L})$ by developing a duality for $\mathbb{ISP}(\mathcal{L})$, where $\mathcal{L}$ is a finite algebra in which underlying lattice is bounded distributive.
Tasks
Published	2019-12-11
URL	https://arxiv.org/abs/1912.12223v1
PDF	https://arxiv.org/pdf/1912.12223v1.pdf
PWC	https://paperswithcode.com/paper/bitopological-duality-for-algebras-of
Repo
Framework

Visualizing RNN States with Predictive Semantic Encodings


Title	Visualizing RNN States with Predictive Semantic Encodings
Authors	Lindsey Sawatzky, Steven Bergner, Fred Popowich
Abstract	Recurrent Neural Networks are an effective and prevalent tool used to model sequential data such as natural language text. However, their deep nature and massive number of parameters pose a challenge for those intending to study precisely how they work. We present a visual technique that gives a high level intuition behind the semantics of the hidden states within Recurrent Neural Networks. This semantic encoding allows for hidden states to be compared throughout the model independent of their internal details. The proposed technique is displayed in a proof of concept visualization tool which is demonstrated to visualize the natural language processing task of language modelling.
Tasks	Language Modelling
Published	2019-08-01
URL	https://arxiv.org/abs/1908.00588v1
PDF	https://arxiv.org/pdf/1908.00588v1.pdf
PWC	https://paperswithcode.com/paper/visualizing-rnn-states-with-predictive
Repo
Framework

How to detect novelty in textual data streams? A comparative study of existing methods


Title	How to detect novelty in textual data streams? A comparative study of existing methods
Authors	Clément Christophe, Julien Velcin, Jairo Cugliari, Philippe Suignard, Manel Boumghar
Abstract	Since datasets with annotation for novelty at the document and/or word level are not easily available, we present a simulation framework that allows us to create different textual datasets in which we control the way novelty occurs. We also present a benchmark of existing methods for novelty detection in textual data streams. We define a few tasks to solve and compare several state-of-the-art methods. The simulation framework allows us to evaluate their performances according to a set of limited scenarios and test their sensitivity to some parameters. Finally, we experiment with the same methods on different kinds of novelty in the New York Times Annotated Dataset.
Tasks
Published	2019-09-11
URL	https://arxiv.org/abs/1909.05099v1
PDF	https://arxiv.org/pdf/1909.05099v1.pdf
PWC	https://paperswithcode.com/paper/how-to-detect-novelty-in-textual-data-streams
Repo
Framework

Knowledge-Grounded Response Generation with Deep Attentional Latent-Variable Model


Title	Knowledge-Grounded Response Generation with Deep Attentional Latent-Variable Model
Authors	Hao-Tong Ye, Kai-Ling Lo, Shang-Yu Su, Yun-Nung Chen
Abstract	End-to-end dialogue generation has achieved promising results without using handcrafted features and attributes specific for each task and corpus. However, one of the fatal drawbacks in such approaches is that they are unable to generate informative utterances, so it limits their usage from some real-world conversational applications. This paper attempts at generating diverse and informative responses with a variational generation model, which contains a joint attention mechanism conditioning on the information from both dialogue contexts and extra knowledge.
Tasks	Dialogue Generation
Published	2019-03-23
URL	http://arxiv.org/abs/1903.09813v1
PDF	http://arxiv.org/pdf/1903.09813v1.pdf
PWC	https://paperswithcode.com/paper/knowledge-grounded-response-generation-with
Repo
Framework

A semi-supervised deep learning algorithm for abnormal EEG identification


Title	A semi-supervised deep learning algorithm for abnormal EEG identification
Authors	Subhrajit Roy, Kiran Kate, Martin Hirzel
Abstract	Systems that can automatically analyze EEG signals can aid neurologists by reducing heavy workload and delays. However, such systems need to be first trained using a labeled dataset. While large corpuses of EEG data exist, a fraction of them are labeled. Hand-labeling data increases workload for the very neurologists we try to aid. This paper proposes a semi-supervised learning workflow that can not only extract meaningful information from large unlabeled EEG datasets but also make predictions with minimal supervision, using labeled datasets as small as 5 examples.
Tasks	EEG
Published	2019-03-19
URL	https://arxiv.org/abs/1903.07822v2
PDF	https://arxiv.org/pdf/1903.07822v2.pdf
PWC	https://paperswithcode.com/paper/a-semi-supervised-deep-learning-algorithm-for
Repo
Framework