Paper Group ANR 1707
From Abstractions to “Natural Languages” for Coordinating Planning Agents. A Unified Bellman Optimality Principle Combining Reward Maximization and Empowerment. Elementos da teoria de aprendizagem de máquina supervisionada. Robust Real-time Pedestrian Detection in Aerial Imagery on Jetson TX2. Learning to Sample: an Active Learning Framework. Autom …
From Abstractions to “Natural Languages” for Coordinating Planning Agents
| Title | From Abstractions to “Natural Languages” for Coordinating Planning Agents |
| Authors | Yu Zhang, Li Wang |
| Abstract | Despite significant advancements in developing autonomous agents, communication between them often relies on a set of pre-specified symbols for a given domain. In this paper, we investigate the automatic construction of these symbols from abstractions to form “natural languages” for such agents. The focus of this initial investigation is on a task planning setting where one agent (the speaker) directly communicates a “plan sketch” to another agent (the listener) to achieve coordination. In contrast to prior work, we view language formation as a fundamental requirement for resolving miscoordination. This view enables us to “compute” a language from the ground up by mapping physical states to symbols, thus reverse engineering the function of languages. Languages that arise from this process are only approximately expressive of the actual plans, meaning that they specify abstractions over the plan space, which is not only theoretically appealing as it provides the desired flexibility to the listener to choose its plan during execution, but also practically useful since it both reduces the communication cost of the speaker and computational cost of the listener. We formulate this language construction problem and show that it is NEXP-complete. An approximate solution is then developed to relate this problem to task planning problems that have efficient off-the-shelf solutions. Finally, we discuss a multi-agent path-finding domain in our evaluation to provide a comprehensive set of results to illustrate the benefits of the constructed languages and their applications. |
| Tasks | Multi-Agent Path Finding |
| Published | 2019-05-01 |
| URL | https://arxiv.org/abs/1905.00517v2 |
| https://arxiv.org/pdf/1905.00517v2.pdf | |
| PWC | https://paperswithcode.com/paper/from-abstractions-to-natural-languages-for |
| Repo | |
| Framework | |
A Unified Bellman Optimality Principle Combining Reward Maximization and Empowerment
| Title | A Unified Bellman Optimality Principle Combining Reward Maximization and Empowerment |
| Authors | Felix Leibfried, Sergio Pascual-Diaz, Jordi Grau-Moya |
| Abstract | Empowerment is an information-theoretic method that can be used to intrinsically motivate learning agents. It attempts to maximize an agent’s control over the environment by encouraging visiting states with a large number of reachable next states. Empowered learning has been shown to lead to complex behaviors, without requiring an explicit reward signal. In this paper, we investigate the use of empowerment in the presence of an extrinsic reward signal. We hypothesize that empowerment can guide reinforcement learning (RL) agents to find good early behavioral solutions by encouraging highly empowered states. We propose a unified Bellman optimality principle for empowered reward maximization. Our empowered reward maximization approach generalizes both Bellman’s optimality principle as well as recent information-theoretical extensions to it. We prove uniqueness of the empowered values and show convergence to the optimal solution. We then apply this idea to develop off-policy actor-critic RL algorithms which we validate in high-dimensional continuous robotics domains (MuJoCo). Our methods demonstrate improved initial and competitive final performance compared to model-free state-of-the-art techniques. |
| Tasks | |
| Published | 2019-07-26 |
| URL | https://arxiv.org/abs/1907.12392v5 |
| https://arxiv.org/pdf/1907.12392v5.pdf | |
| PWC | https://paperswithcode.com/paper/a-unified-bellman-optimality-principle |
| Repo | |
| Framework | |
Elementos da teoria de aprendizagem de máquina supervisionada
| Title | Elementos da teoria de aprendizagem de máquina supervisionada |
| Authors | Vladimir G. Pestov |
| Abstract | This is a set of lecture notes for an introductory course (advanced undergaduates or the 1st graduate course) on foundations of supervised machine learning (in Portuguese). The topics include: the geometry of the Hamming cube, concentration of measure, shattering and VC dimension, Glivenko-Cantelli classes, PAC learnability, universal consistency and the k-NN classifier in metric spaces, dimensionality reduction, universal approximation, sample compression. There are appendices on metric and normed spaces, measure theory, etc., making the notes self-contained. Este 'e um conjunto de notas de aula para um curso introdut'orio (curso de gradua\c{c}~ao avan\c{c}ado ou o 1o curso de p'os) sobre fundamentos da aprendizagem de m'aquina supervisionada (em Portugu^es). Os t'opicos incluem: a geometria do cubo de Hamming, concentra\c{c}~ao de medida, fragmenta\c{c}~ao e dimens~ao de Vapnik-Chervonenkis, classes de Glivenko-Cantelli, aprendizabilidade PAC, consist^encia universal e o classificador k-NN em espa\c{c}os m'etricos, redu\c{c}~ao de dimensionalidade, aproxima\c{c}~ao universal, compress~ao amostral. H'a ap^endices sobre espa\c{c}os m'etricos e normados, teoria de medida, etc., tornando as notas autosuficientes. |
| Tasks | Dimensionality Reduction |
| Published | 2019-10-06 |
| URL | https://arxiv.org/abs/1910.06820v1 |
| https://arxiv.org/pdf/1910.06820v1.pdf | |
| PWC | https://paperswithcode.com/paper/elementos-da-teoria-de-aprendizagem-de |
| Repo | |
| Framework | |
Robust Real-time Pedestrian Detection in Aerial Imagery on Jetson TX2
| Title | Robust Real-time Pedestrian Detection in Aerial Imagery on Jetson TX2 |
| Authors | Mohamed Afifi, Yara Ali, Karim Amer, Mahmoud Shaker, Mohamed ElHelw |
| Abstract | Detection of pedestrians in aerial imagery captured by drones has many applications including intersection monitoring, patrolling, and surveillance, to name a few. However, the problem is involved due to continuouslychanging camera viewpoint and object appearance as well as the need for lightweight algorithms to run on on-board embedded systems. To address this issue, the paper proposes a framework for pedestrian detection in videos based on the YOLO object detection network [6] while having a high throughput of more than 5 FPS on the Jetson TX2 embedded board. The framework exploits deep learning for robust operation and uses a pre-trained model without the need for any additional training which makes it flexible to apply on different setups with minimum amount of tuning. The method achieves ~81 mAP when applied on a sample video from the Embedded Real-Time Inference (ERTI) Challenge where pedestrians are monitored by a UAV. |
| Tasks | Object Detection, Pedestrian Detection |
| Published | 2019-05-16 |
| URL | https://arxiv.org/abs/1905.06653v1 |
| https://arxiv.org/pdf/1905.06653v1.pdf | |
| PWC | https://paperswithcode.com/paper/robust-real-time-pedestrian-detection-in |
| Repo | |
| Framework | |
Learning to Sample: an Active Learning Framework
| Title | Learning to Sample: an Active Learning Framework |
| Authors | Jingyu Shao, Qing Wang, Fangbing Liu |
| Abstract | Meta-learning algorithms for active learning are emerging as a promising paradigm for learning the ``best’’ active learning strategy. However, current learning-based active learning approaches still require sufficient training data so as to generalize meta-learning models for active learning. This is contrary to the nature of active learning which typically starts with a small number of labeled samples. The unavailability of large amounts of labeled samples for training meta-learning models would inevitably lead to poor performance (e.g., instabilities and overfitting). In our paper, we tackle these issues by proposing a novel learning-based active learning framework, called Learning To Sample (LTS). This framework has two key components: a sampling model and a boosting model, which can mutually learn from each other in iterations to improve the performance of each other. Within this framework, the sampling model incorporates uncertainty sampling and diversity sampling into a unified process for optimization, enabling us to actively select the most representative and informative samples based on an optimized integration of uncertainty and diversity. To evaluate the effectiveness of the LTS framework, we have conducted extensive experiments on three different classification tasks: image classification, salary level prediction, and entity resolution. The experimental results show that our LTS framework significantly outperforms all the baselines when the label budget is limited, especially for datasets with highly imbalanced classes. In addition to this, our LTS framework can effectively tackle the cold start problem occurring in many existing active learning approaches. | |
| Tasks | Active Learning, Entity Resolution, Image Classification, Meta-Learning |
| Published | 2019-09-09 |
| URL | https://arxiv.org/abs/1909.03585v1 |
| https://arxiv.org/pdf/1909.03585v1.pdf | |
| PWC | https://paperswithcode.com/paper/learning-to-sample-an-active-learning |
| Repo | |
| Framework | |
Automatic assessment of spoken language proficiency of non-native children
| Title | Automatic assessment of spoken language proficiency of non-native children |
| Authors | Roberto Gretter, Katharina Allgaier, Svetlana Tchistiakova, Daniele Falavigna |
| Abstract | This paper describes technology developed to automatically grade Italian students (ages 9-16) on their English and German spoken language proficiency. The students’ spoken answers are first transcribed by an automatic speech recognition (ASR) system and then scored using a feedforward neural network (NN) that processes features extracted from the automatic transcriptions. In-domain acoustic models, employing deep neural networks (DNNs), are derived by adapting the parameters of an original out of domain DNN. |
| Tasks | Speech Recognition |
| Published | 2019-03-15 |
| URL | http://arxiv.org/abs/1903.06409v1 |
| http://arxiv.org/pdf/1903.06409v1.pdf | |
| PWC | https://paperswithcode.com/paper/automatic-assessment-of-spoken-language |
| Repo | |
| Framework | |
Adversarial Neural Pruning with Latent Vulnerability Suppression
| Title | Adversarial Neural Pruning with Latent Vulnerability Suppression |
| Authors | Divyam Madaan, Jinwoo Shin, Sung Ju Hwang |
| Abstract | Despite the remarkable performance of deep neural networks on various computer vision tasks, they are known to be highly susceptible to adversarial perturbations which makes it difficult to deploy them in real-world safety-critical applications. In this paper, we conjecture that the main cause of this adversarial vulnerability is the distortion in the latent feature space, and provide methods to effectively suppress them. Specifically, we define \textbf{vulnerability} for each latent feature and then propose a new loss for adversarial learning, \textbf{Vulnerability Suppression (VS)} loss, that aims to minimize the feature-level vulnerability during training. We further propose a Bayesian framework to prune features with high vulnerability, in order to reduce both vulnerability and loss on adversarial samples. We validate our \textbf{Adversarial Neural Pruning (ANP)} method on multiple benchmark datasets, on which it not only obtains state-of-the-art adversarial robustness but also improves the performance on clean examples, using only a fraction of the parameters used by the full network. Further qualitative analysis suggests that the improvements actually come from the suppression of feature-level vulnerability. |
| Tasks | |
| Published | 2019-08-12 |
| URL | https://arxiv.org/abs/1908.04355v3 |
| https://arxiv.org/pdf/1908.04355v3.pdf | |
| PWC | https://paperswithcode.com/paper/adversarial-neural-pruning |
| Repo | |
| Framework | |
Stochastic Variance Reduced Primal Dual Algorithms for Empirical Composition Optimization
| Title | Stochastic Variance Reduced Primal Dual Algorithms for Empirical Composition Optimization |
| Authors | Adithya M. Devraj, Jianshu Chen |
| Abstract | We consider a generic empirical composition optimization problem, where there are empirical averages present both outside and inside nonlinear loss functions. Such a problem is of interest in various machine learning applications, and cannot be directly solved by standard methods such as stochastic gradient descent. We take a novel approach to solving this problem by reformulating the original minimization objective into an equivalent min-max objective, which brings out all the empirical averages that are originally inside the nonlinear loss functions. We exploit the rich structures of the reformulated problem and develop a stochastic primal-dual algorithm, SVRPDA-I, to solve the problem efficiently. We carry out extensive theoretical analysis of the proposed algorithm, obtaining the convergence rate, the computation complexity and the storage complexity. In particular, the algorithm is shown to converge at a linear rate when the problem is strongly convex. Moreover, we also develop an approximate version of the algorithm, named SVRPDA-II, which further reduces the memory requirement. Finally, we evaluate our proposed algorithms on several real-world benchmarks, and experimental results show that the proposed algorithms significantly outperform existing techniques. |
| Tasks | |
| Published | 2019-07-22 |
| URL | https://arxiv.org/abs/1907.09150v2 |
| https://arxiv.org/pdf/1907.09150v2.pdf | |
| PWC | https://paperswithcode.com/paper/stochastic-variance-reduced-primal-dual |
| Repo | |
| Framework | |
Deep Fusion: An Attention Guided Factorized Bilinear Pooling for Audio-video Emotion Recognition
| Title | Deep Fusion: An Attention Guided Factorized Bilinear Pooling for Audio-video Emotion Recognition |
| Authors | Yuanyuan Zhang, Zi-Rui Wang, Jun Du |
| Abstract | Automatic emotion recognition (AER) is a challenging task due to the abstract concept and multiple expressions of emotion. Although there is no consensus on a definition, human emotional states usually can be apperceived by auditory and visual systems. Inspired by this cognitive process in human beings, it’s natural to simultaneously utilize audio and visual information in AER. However, most traditional fusion approaches only build a linear paradigm, such as feature concatenation and multi-system fusion, which hardly captures complex association between audio and video. In this paper, we introduce factorized bilinear pooling (FBP) to deeply integrate the features of audio and video. Specifically, the features are selected through the embedded attention mechanism from respective modalities to obtain the emotion-related regions. The whole pipeline can be completed in a neural network. Validated on the AFEW database of the audio-video sub-challenge in EmotiW2018, the proposed approach achieves an accuracy of 62.48%, outperforming the state-of-the-art result. |
| Tasks | Emotion Recognition, Video Emotion Recognition |
| Published | 2019-01-15 |
| URL | http://arxiv.org/abs/1901.04889v1 |
| http://arxiv.org/pdf/1901.04889v1.pdf | |
| PWC | https://paperswithcode.com/paper/deep-fusion-an-attention-guided-factorized |
| Repo | |
| Framework | |
Finite sample properties of parametric MMD estimation: robustness to misspecification and dependence
| Title | Finite sample properties of parametric MMD estimation: robustness to misspecification and dependence |
| Authors | Badr-Eddine Chérief-Abdellatif, Pierre Alquier |
| Abstract | Many works in statistics aim at designing a universal estimation procedure. This question is of major interest, in particular because it leads to robust estimators, a very hot topic in statistics and machine learning. In this paper, we tackle the problem of universal estimation using a minimum distance estimator presented in Briol et al. (2019) based on the Maximum Mean Discrepancy. We show that the estimator is robust to both dependence and to the presence of outliers in the dataset. We also highlight the connections that may exist with minimum distance estimators using L2-distance. Finally, we provide a theoretical study of the stochastic gradient descent algorithm used to compute the estimator, and we support our findings with numerical simulations. |
| Tasks | |
| Published | 2019-12-12 |
| URL | https://arxiv.org/abs/1912.05737v3 |
| https://arxiv.org/pdf/1912.05737v3.pdf | |
| PWC | https://paperswithcode.com/paper/finite-sample-properties-of-parametric-mmd |
| Repo | |
| Framework | |
Robust Guarantees for Perception-Based Control
| Title | Robust Guarantees for Perception-Based Control |
| Authors | Sarah Dean, Nikolai Matni, Benjamin Recht, Vickie Ye |
| Abstract | Motivated by vision-based control of autonomous vehicles, we consider the problem of controlling a known linear dynamical system for which partial state information, such as vehicle position, is extracted from complex and nonlinear data, such as a camera image. Our approach is to use a learned perception map that predicts some linear function of the state and to design a corresponding safe set and robust controller for the closed loop system with this sensing scheme. We show that under suitable smoothness assumptions on both the perception map and the generative model relating state to complex and nonlinear data, parameters of the safe set can be learned via appropriately dense sampling of the state space. We then prove that the resulting perception-control loop has favorable generalization properties. We illustrate the usefulness of our approach on a synthetic example and on the self-driving car simulation platform CARLA. |
| Tasks | Autonomous Vehicles |
| Published | 2019-07-08 |
| URL | https://arxiv.org/abs/1907.03680v2 |
| https://arxiv.org/pdf/1907.03680v2.pdf | |
| PWC | https://paperswithcode.com/paper/robust-guarantees-for-perception-based |
| Repo | |
| Framework | |
Joint Neural Collaborative Filtering for Recommender Systems
| Title | Joint Neural Collaborative Filtering for Recommender Systems |
| Authors | Wanyu Chen, Fei Cai, Honghui Chen, Maarten de Rijke |
| Abstract | We propose a J-NCF method for recommender systems. The J-NCF model applies a joint neural network that couples deep feature learning and deep interaction modeling with a rating matrix. Deep feature learning extracts feature representations of users and items with a deep learning architecture based on a user-item rating matrix. Deep interaction modeling captures non-linear user-item interactions with a deep neural network using the feature representations generated by the deep feature learning process as input. J-NCF enables the deep feature learning and deep interaction modeling processes to optimize each other through joint training, which leads to improved recommendation performance. In addition, we design a new loss function for optimization, which takes both implicit and explicit feedback, point-wise and pair-wise loss into account. Experiments on several real-word datasets show significant improvements of J-NCF over state-of-the-art methods, with improvements of up to 8.24% on the MovieLens 100K dataset, 10.81% on the MovieLens 1M dataset, and 10.21% on the Amazon Movies dataset in terms of HR@10. NDCG@10 improvements are 12.42%, 14.24% and 15.06%, respectively. We also conduct experiments to evaluate the scalability and sensitivity of J-NCF. Our experiments show that the J-NCF model has a competitive recommendation performance with inactive users and different degrees of data sparsity when compared to state-of-the-art baselines. |
| Tasks | Recommendation Systems |
| Published | 2019-07-08 |
| URL | https://arxiv.org/abs/1907.03459v2 |
| https://arxiv.org/pdf/1907.03459v2.pdf | |
| PWC | https://paperswithcode.com/paper/joint-neural-collaborative-filtering-for |
| Repo | |
| Framework | |
Optimize TSK Fuzzy Systems for Regression Problems: Mini-Batch Gradient Descent with Regularization, DropRule and AdaBound (MBGD-RDA)
| Title | Optimize TSK Fuzzy Systems for Regression Problems: Mini-Batch Gradient Descent with Regularization, DropRule and AdaBound (MBGD-RDA) |
| Authors | Dongrui Wu, Ye Yuan, Yihua Tan |
| Abstract | Takagi-Sugeno-Kang (TSK) fuzzy systems are very useful machine learning models for regression problems. However, to our knowledge, there has not existed an efficient and effective training algorithm that ensures their generalization performance, and also enables them to deal with big data. Inspired by the connections between TSK fuzzy systems and neural networks, we extend three powerful neural network optimization techniques, i.e., mini-batch gradient descent, regularization, and AdaBound, to TSK fuzzy systems, and also propose three novel techniques (DropRule, DropMF, and DropMembership) specifically for training TSK fuzzy systems. Our final algorithm, mini-batch gradient descent with regularization, DropRule and AdaBound (MBGD-RDA), can achieve fast convergence in training TSK fuzzy systems, and also superior generalization performance in testing. It can be used for training TSK fuzzy systems on datasets of any size; however, it is particularly useful for big datasets, on which currently no other efficient training algorithms exist. |
| Tasks | |
| Published | 2019-03-26 |
| URL | https://arxiv.org/abs/1903.10951v4 |
| https://arxiv.org/pdf/1903.10951v4.pdf | |
| PWC | https://paperswithcode.com/paper/optimize-tsk-fuzzy-systems-for-big-data |
| Repo | |
| Framework | |
Asymptotic Unbiasedness of the Permutation Importance Measure in Random Forest Models
| Title | Asymptotic Unbiasedness of the Permutation Importance Measure in Random Forest Models |
| Authors | Burim Ramosaj, Markus Pauly |
| Abstract | Variable selection in sparse regression models is an important task as applications ranging from biomedical research to econometrics have shown. Especially for higher dimensional regression problems, for which the link function between response and covariates cannot be directly detected, the selection of informative variables is challenging. Under these circumstances, the Random Forest method is a helpful tool to predict new outcomes while delivering measures for variable selection. One common approach is the usage of the permutation importance. Due to its intuitive idea and flexible usage, it is important to explore circumstances, for which the permutation importance based on Random Forest correctly indicates informative covariates. Regarding the latter, we deliver theoretical guarantees for the validity of the permutation importance measure under specific assumptions and prove its (asymptotic) unbiasedness. An extensive simulation study verifies our findings. |
| Tasks | |
| Published | 2019-12-05 |
| URL | https://arxiv.org/abs/1912.03306v1 |
| https://arxiv.org/pdf/1912.03306v1.pdf | |
| PWC | https://paperswithcode.com/paper/asymptotic-unbiasedness-of-the-permutation |
| Repo | |
| Framework | |
CTNN: Corticothalamic-inspired neural network
| Title | CTNN: Corticothalamic-inspired neural network |
| Authors | Leendert A Remmelzwaal, Amit K Mishra, George F R Ellis |
| Abstract | Sensory predictions by the brain in all modalities take place as a result of bottom-up and top-down connections both in the neocortex and between the neocortex and the thalamus. The bottom-up connections in the cortex are responsible for learning, pattern recognition, and object classification, and have been widely modelled using artificial neural networks (ANNs). Here, we present a neural network architecture modelled on the top-down corticothalamic connections and the behaviour of the thalamus: a corticothalamic neural network (CTNN), consisting of an auto-encoder connected to a difference engine with a threshold. We demonstrate that the CTNN is input agnostic, multi-modal, robust during partial occlusion of one or more sensory inputs, and has significantly higher processing efficiency than other predictive coding models, proportional to the number of sequentially similar inputs in a sequence. This increased efficiency could be highly significant in more complex implementations of this architecture, where the predictive nature of the cortex will allow most of the incoming data to be discarded. |
| Tasks | Object Classification |
| Published | 2019-10-28 |
| URL | https://arxiv.org/abs/1910.12492v2 |
| https://arxiv.org/pdf/1910.12492v2.pdf | |
| PWC | https://paperswithcode.com/paper/ctnn-corticothalamic-inspired-neural-network |
| Repo | |
| Framework | |
