A key challenge for reinforcement learning (RL) consists of learning in environments with sparse extrinsic rewards. In contrast to current RL methods, humans are able to learn new skills with little or no reward by using various forms of intrinsic motivation. We propose AMIGo, a novel agent incorporating a goal-generating teacher that proposes Adversarially Motivated Intrinsic Goals to train a goal-conditioned "student" policy in the absence of (or alongside) environment reward. Specifically, through a simple but effective "constructively adversarial" objective, the teacher learns to propose increasingly challenging---yet achievable---goals that allow the student to learn general skills for acting in a new environment, independent of the task to be solved. We show that our method generates a natural curriculum of self-proposed goals which ultimately allows the agent to solve challenging procedurally-generated tasks where other forms of intrinsic motivation and state-of-the-art RL methods fail.

Knowledge Graph Completion (KGC) has been proposed to improve Knowledge Graphs by filling in missing connections via link prediction or relation extraction. One of the main difficulties for KGC is a low resource problem. Previous approaches assume sufficient training triples to learn versatile vectors for entities and relations, or a satisfactory number of labeled sentences to train a competent relation extraction model. However, low resource relations are very common in KGs, and those newly added relations often do not have many known samples for training. In this work, we aim at predicting new facts under a challenging setting where only limited training instances are available. We propose a general framework called Weighted Relation Adversarial Network, which utilizes an adversarial procedure to help adapt knowledge/features learned from high resource relations to different but related low resource relations. Specifically, the framework takes advantage of a relation discriminator to distinguish between samples from different relations, and help learn relation-invariant features more transferable from source relations to target relations. Experimental results show that the proposed approach outperforms previous methods regarding low resource settings for both link prediction and relation extraction.

Graph Convolutional Network (GCN) has achieved extraordinary success in learning effective task-specific representations of nodes in graphs. However, regarding Heterogeneous Information Network (HIN), existing HIN-oriented GCN methods still suffer from two deficiencies: (1) they cannot flexibly explore all possible meta-paths and extract the most useful ones for a target object, which hinders both effectiveness and interpretability; (2) they often need to generate intermediate meta-path based dense graphs, which leads to high computational complexity. To address the above issues, we propose an interpretable and efficient Heterogeneous Graph Convolutional Network (ie-HGCN) to learn the representations of objects in HINs. It is designed as a hierarchical aggregation architecture, i.e., object-level aggregation first, followed by type-level aggregation. The novel architecture can automatically extract useful meta-paths for each object from all possible meta-paths (within a length limit), which brings good model interpretability. It can also reduce the computational cost by avoiding intermediate HIN transformation and neighborhood attention. We provide theoretical analysis about the proposed ie-HGCN in terms of evaluating the usefulness of all possible meta-paths, its connection to the spectral graph convolution on HINs, and its quasi-linear time complexity. Extensive experiments on three real network datasets demonstrate the superiority of ie-HGCN over the state-of-the-art methods.

Increasing the scale of reinforcement learning experiments has allowed researchers to achieve unprecedented results in both training sophisticated agents for video games, and in sim-to-real transfer for robotics. Typically such experiments rely on large distributed systems and require expensive hardware setups, limiting wider access to this exciting area of research. In this work we aim to solve this problem by optimizing the efficiency and resource utilization of reinforcement learning algorithms instead of relying on distributed computation. We present the "Sample Factory", a high-throughput training system optimized for a single-machine setting. Our architecture combines a highly efficient, asynchronous, GPU-based sampler with off-policy correction techniques, allowing us to achieve throughput higher than $10^5$ environment frames/second on non-trivial control problems in 3D without sacrificing sample efficiency. We extend Sample Factory to support self-play and population-based training and apply these techniques to train highly capable agents for a multiplayer first-person shooter game. The source code is available at https://github.com/alex-petrenko/sample-factory

Machine learning models have been successfully applied to a wide range of applications including computer vision, natural language processing, and speech recognition. A successful implementation of these models however, usually relies on deep neural networks (DNNs) which are treated as opaque black-box systems due to their incomprehensible complexity and intricate internal mechanism. In this work, we present a novel algorithm for explaining the predictions of a DNN using adversarial machine learning. Our approach identifies the relative importance of input features in relation to the predictions based on the behavior of an adversarial attack on the DNN. Our algorithm has the advantage of being fast, consistent, and easy to implement and interpret. We present our detailed analysis that demonstrates how the behavior of an adversarial attack, given a DNN and a task, stays consistent for any input test data point proving the generality of our approach. Our analysis enables us to produce consistent and efficient explanations. We illustrate the effectiveness of our approach by conducting experiments using a variety of DNNs, tasks, and datasets. Finally, we compare our work with other well-known techniques in the current literature.

Nowadays, deep neural networks are widely used in mission critical systems such as healthcare, self-driving vehicles, and military which have direct impact on human lives. However, the black-box nature of deep neural networks challenges its use in mission critical applications, raising ethical and judicial concerns inducing lack of trust. Explainable Artificial Intelligence (XAI) is a field of Artificial Intelligence (AI) that promotes a set of tools, techniques, and algorithms that can generate high-quality interpretable, intuitive, human-understandable explanations of AI decisions. In addition to providing a holistic view of the current XAI landscape in deep learning, this paper provides mathematical summaries of seminal work. We start by proposing a taxonomy and categorizing the XAI techniques based on their scope of explanations, methodology behind the algorithms, and explanation level or usage which helps build trustworthy, interpretable, and self-explanatory deep learning models. We then describe the main principles used in XAI research and present the historical timeline for landmark studies in XAI from 2007 to 2020. After explaining each category of algorithms and approaches in detail, we then evaluate the explanation maps generated by eight XAI algorithms on image data, discuss the limitations of this approach, and provide potential future directions to improve XAI evaluation.

Multi-step (also called n-step) methods in reinforcement learning (RL) have been shown to be more efficient than the 1-step method due to faster propagation of the reward signal, both theoretically and empirically, in tasks exploiting tabular representation of the value-function. Recently, research in Deep Reinforcement Learning (DRL) also shows that multi-step methods improve learning speed and final performance in applications where the value-function and policy are represented with deep neural networks. However, there is a lack of understanding about what is actually contributing to the boost of performance. In this work, we analyze the effect of multi-step methods on alleviating the overestimation problem in DRL, where multi-step experiences are sampled from a replay buffer. Specifically building on top of Deep Deterministic Policy Gradient (DDPG), we propose Multi-step DDPG (MDDPG), where different step sizes are manually set, and its variant called Mixed Multi-step DDPG (MMDDPG) where an average over different multi-step backups is used as update target of Q-value function. Empirically, we show that both MDDPG and MMDDPG are significantly less affected by the overestimation problem than DDPG with 1-step backup, which consequently results in better final performance and learning speed. We also discuss the advantages and disadvantages of different ways to do multi-step expansion in order to reduce approximation error, and expose the tradeoff between overestimation and underestimation that underlies offline multi-step methods. Finally, we compare the computational resource needs of Twin Delayed Deep Deterministic Policy Gradient (TD3), a state-of-art algorithm proposed to address overestimation in actor-critic methods, and our proposed methods, since they show comparable final performance and learning speed.

We introduce MotioNet, a deep neural network that directly reconstructs the motion of a 3D human skeleton from monocular video.While previous methods rely on either rigging or inverse kinematics (IK) to associate a consistent skeleton with temporally coherent joint rotations, our method is the first data-driven approach that directly outputs a kinematic skeleton, which is a complete, commonly used, motion representation. At the crux of our approach lies a deep neural network with embedded kinematic priors, which decomposes sequences of 2D joint positions into two separate attributes: a single, symmetric, skeleton, encoded by bone lengths, and a sequence of 3D joint rotations associated with global root positions and foot contact labels. These attributes are fed into an integrated forward kinematics (FK) layer that outputs 3D positions, which are compared to a ground truth. In addition, an adversarial loss is applied to the velocities of the recovered rotations, to ensure that they lie on the manifold of natural joint rotations. The key advantage of our approach is that it learns to infer natural joint rotations directly from the training data, rather than assuming an underlying model, or inferring them from joint positions using a data-agnostic IK solver. We show that enforcing a single consistent skeleton along with temporally coherent joint rotations constrains the solution space, leading to a more robust handling of self-occlusions and depth ambiguities.

The use of low-cost sensors in air quality monitoring networks is still a much-debated topic among practitioners: they are much cheaper than traditional air quality monitoring stations set up by public authorities (a few hundred dollars compared to a few dozens of thousand dollars) at the cost of a lower accuracy and robustness. This paper presents a case study of using low-cost sensors measurements in an air quality prediction engine. The engine predicts jointly PM2.5 and PM10 concentrations in the United States at a very high resolution in the range of a few dozens of meters. It is fed with the measurements provided by official air quality monitoring stations, the measurements provided by a network of more than 4000 low-cost sensors across the country, and traffic estimates. We show that the use of low-cost sensors' measurements improves the engine's accuracy very significantly. In particular, we derive a strong link between the density of low-cost sensors and the predictions' accuracy: the more low-cost sensors are in an area, the more accurate are the predictions. As an illustration, in areas with the highest density of low-cost sensors, the low-cost sensors' measurements bring a 25% and 15% improvement in PM2.5 and PM10 predictions' accuracy respectively. An other strong conclusion is that in some areas with a high density of low-cost sensors, the engine performs better when fed with low-cost sensors' measurements only than when fed with official monitoring stations' measurements only: this suggests that an air quality monitoring network composed of low-cost sensors is effective in monitoring air quality. This is a very important result, as such a monitoring network is much cheaper to set up.

Learning interpretable and disentangled representations is a crucial yet challenging task in representation learning. In this work, we focus on semi-supervised disentanglement learning and extend work by Locatello et al. (2019) by introducing another source of supervision that we denote as label replacement. Specifically, during training, we replace the inferred representation associated with a data point with its ground-truth representation whenever it is available. Our extension is theoretically inspired by our proposed general framework of semi-supervised disentanglement learning in the context of VAEs which naturally motivates the supervised terms commonly used in existing semi-supervised VAEs (but not for disentanglement learning). Extensive experiments on synthetic and real datasets demonstrate both quantitatively and qualitatively the ability of our extension to significantly and consistently improve disentanglement with very limited supervision.