Alan R. Wagner

Ali Ayub, Alan R. Wagner

For most real-world applications, robots need to adapt and learn continually with limited data in their environments. In this paper, we consider the problem of Few-Shot class Incremental Learning (FSIL), in which an AI agent is required to learn incrementally from a few data samples without forgetting the data it has previously learned. To solve this problem, we present a novel framework inspired by theories of concept learning in the hippocampus and the neocortex. Our framework represents object classes in the form of sets of clusters and stores them in memory. The framework replays data generated by the clusters of the old classes, to avoid forgetting when learning new classes. Our approach is evaluated on two object classification datasets resulting in state-of-the-art (SOTA) performance for class-incremental learning and FSIL. We also evaluate our framework for FSIL on a robot demonstrating that the robot can continually learn to classify a large set of household objects with limited human assistance.

Alan R. Wagner, Colin Holbrook, Daniel Holman et al.

This paper describes our recent effort to use virtual reality to simulate threatening emergency evacuation scenarios in which a robot guides a person to an exit. Our prior work has demonstrated that people will follow a robot's guidance, even when the robot is faulty, during an emergency evacuation. Yet, because physical in-person emergency evacuation experiments are difficult and costly to conduct and because we would like to evaluate many different factors, we are motivated to develop a system that immerses people in the simulation environment to encourage genuine subject reactions. We are working to complete experiments verifying the validity of our approach.

Christopher A. McClurg, Alan R. Wagner

Virtual reality (VR) has emerged as a powerful tool for evaluating school security measures in high-risk scenarios such as school shootings, offering experimental control and high behavioral fidelity. However, assessing new interventions in VR requires recruiting new participant cohorts for each condition, making large-scale or iterative evaluation difficult. These limitations are especially restrictive when attempting to learn effective intervention strategies, which typically require many training episodes. To address this challenge, we develop a data-driven discrete-event simulator (DES) that models shooter movement and in-region actions as stochastic processes learned from participant behavior in VR studies. We use the simulator to examine the impact of a robot-based shooter intervention strategy. Once shown to reproduce key empirical patterns, the DES enables scalable evaluation and learning of intervention strategies that are infeasible to train directly with human subjects. Overall, this work demonstrates a high-to-mid fidelity simulation workflow that provides a scalable surrogate for developing and evaluating autonomous school-security interventions.

Christopher McClurg, Ali Ayub, Harsh Tyagi et al.

For real-world applications, robots will need to continually learn in their environments through limited interactions with their users. Toward this, previous works in few-shot class incremental learning (FSCIL) and active class selection (ACS) have achieved promising results but were tested in constrained setups. Therefore, in this paper, we combine ideas from FSCIL and ACS to develop a novel framework that can allow an autonomous agent to continually learn new objects by asking its users to label only a few of the most informative objects in the environment. To this end, we build on a state-of-the-art (SOTA) FSCIL model and extend it with techniques from ACS literature. We term this model Few-shot Incremental Active class SeleCtiOn (FIASco). We further integrate a potential field-based navigation technique with our model to develop a complete framework that can allow an agent to process and reason on its sensory data through the FIASco model, navigate towards the most informative object in the environment, gather data about the object through its sensors and incrementally update the FIASco model. Experimental results on a simulated agent and a real robot show the significance of our approach for long-term real-world robotics applications.

Ali Ayub, Huiqing Hu, Guangwei Zhou et al.

Robots may soon play a role in higher education by augmenting learning environments and managing interactions between instructors and learners. Little, however, is known about how the presence of robots in the learning environment will influence academic integrity. This study therefore investigates if and how college students cheat while engaged in a collaborative sorting task with a robot. We employed a 2x2 factorial design to examine the effects of cheating exposure (exposure to cheating or no exposure) and task clarity (clear or vague rules) on college student cheating behaviors while interacting with a robot. Our study finds that prior exposure to cheating on the task significantly increases the likelihood of cheating. Yet, the tendency to cheat was not impacted by the clarity of the task rules. These results suggest that normative behavior by classmates may strongly influence the decision to cheat while engaged in an instructional experience with a robot.

Kasra Mokhtari, Alan R. Wagner

Risk is traditionally described as the expected likelihood of an undesirable outcome, such as collisions for autonomous vehicles. Accurately predicting risk or potentially risky situations is critical for the safe operation of autonomous vehicles. In our previous work, we showed that risk could be characterized by two components: 1) the probability of an undesirable outcome and 2) an estimate of how undesirable the outcome is (loss). This paper is an extension to our previous work. In this paper, using our trained deep reinforcement learning model for navigating around crowds, we developed a risk-based decision-making framework for the autonomous vehicle that integrates the high-level risk-based path planning with the reinforcement learning-based low-level control. We evaluated our method in a high-fidelity simulation such as CARLA. This work can improve safety by allowing an autonomous vehicle to one day avoid and react to risky situations.

Kasra Mokhtari, Alan R. Wagner

We propose a safe DRL approach for autonomous vehicle (AV) navigation through crowds of pedestrians while making a left turn at an unsignalized intersection. Our method uses two long-short term memory (LSTM) models that are trained to generate the perceived state of the environment and the future trajectories of pedestrians given noisy observations of their movement. A future collision prediction algorithm based on the future trajectories of the ego vehicle and pedestrians is used to mask unsafe actions if the system predicts a collision. The performance of our approach is evaluated in two experiments using the high-fidelity CARLA simulation environment. The first experiment tests the performance of our method at intersections that are similar to the training intersection and the second experiment tests our method at intersections with a different topology. For both experiments, our methods do not result in a collision with a pedestrian while still navigating the intersection at a reasonable speed.

Kasra Mokhtari, Alan R. Wagner

Prior research has extensively explored Autonomous Vehicle (AV) navigation in the presence of other vehicles, however, navigation among pedestrians, who are the most vulnerable element in urban environments, has been less examined. This paper explores AV navigation in crowded, unsignalized intersections. We compare the performance of different deep reinforcement learning methods trained on our reward function and state representation. The performance of these methods and a standard rule-based approach were evaluated in two ways, first at the unsignalized intersection on which the methods were trained, and secondly at an unknown unsignalized intersection with a different topology. For both scenarios, the rule-based method achieves less than 40\% collision-free episodes, whereas our methods result in a performance of approximately 100\%. Of the three methods used, DDQN/PER outperforms the other two methods while it also shows the smallest average intersection crossing time, the greatest average speed, and the greatest distance from the closest pedestrian.

Ali Ayub, Alan R. Wagner

Deep learning has achieved remarkable success in object recognition tasks through the availability of large scale datasets like ImageNet. However, deep learning systems suffer from catastrophic forgetting when learning incrementally without replaying old data. For real-world applications, robots also need to incrementally learn new objects. Further, since robots have limited human assistance available, they must learn from only a few examples. However, very few object recognition datasets and benchmarks exist to test incremental learning capability for robotic vision. Further, there is no dataset or benchmark specifically designed for incremental object learning from a few examples. To fill this gap, we present a new dataset termed F-SIOL-310 (Few-Shot Incremental Object Learning) which is specifically captured for testing few-shot incremental object learning capability for robotic vision. We also provide benchmarks and evaluations of 8 incremental learning algorithms on F-SIOL-310 for future comparisons. Our results demonstrate that the few-shot incremental object learning problem for robotic vision is far from being solved.

Ali Ayub, Alan R. Wagner

Children learn continually by asking questions about the concepts they are most curious about. With robots becoming an integral part of our society, they must also learn unknown concepts continually by asking humans questions. The paper analyzes a recent state-of-the-art approach for continual learning. The paper further develops a self-supervised technique to find most of the uncertain objects in an environment by utilizing the cluster representation of the previously learned classes. We test our approach on a benchmark dataset for continual learning on robots. Our results show that our curiosity-driven continual learning approach beats random sampling and softmax-based uncertainty sampling in terms of classification accuracy and the total number of classes learned.

Ali Ayub, Alan R. Wagner

For many real-world robotics applications, robots need to continually adapt and learn new concepts. Further, robots need to learn through limited data because of scarcity of labeled data in the real-world environments. To this end, my research focuses on developing robots that continually learn in dynamic unseen environments/scenarios, learn from limited human supervision, remember previously learned knowledge and use that knowledge to learn new concepts. I develop machine learning models that not only produce State-of-the-results on benchmark datasets but also allow robots to learn new objects and scenes in unconstrained environments which lead to a variety of novel robotics applications.

Ali Ayub, Alan R. Wagner

The two main impediments to continual learning are catastrophic forgetting and memory limitations on the storage of data. To cope with these challenges, we propose a novel, cognitively-inspired approach which trains autoencoders with Neural Style Transfer to encode and store images. During training on a new task, reconstructed images from encoded episodes are replayed in order to avoid catastrophic forgetting. The loss function for the reconstructed images is weighted to reduce its effect during classifier training to cope with image degradation. When the system runs out of memory the encoded episodes are converted into centroids and covariance matrices, which are used to generate pseudo-images during classifier training, keeping classifier performance stable while using less memory. Our approach increases classification accuracy by 13-17% over state-of-the-art methods on benchmark datasets, while requiring 78% less storage space.

Ali Ayub, Alan R. Wagner

Social norms support coordination and cooperation in society. With social robots becoming increasingly involved in our society, they also need to follow the social norms of the society. This paper presents a computational framework for learning contexts and the social norms present in a context in an online manner on a robot. The paper utilizes a recent state-of-the-art approach for incremental learning and adapts it for online learning of scenes (contexts). The paper further utilizes Dempster-Schafer theory to model context-specific norms. After learning the scenes (contexts), we use active learning to learn related norms. We test our approach on the Pepper robot by taking it through different scene locations. Our results show that Pepper can learn different scenes and related norms simply by communicating with a human partner in an online manner.

Ali Ayub, Alan R. Wagner

For many applications, robots will need to be incrementally trained to recognize the specific objects needed for an application. This paper presents a practical system for incrementally training a robot to recognize different object categories using only a small set of visual examples provided by a human. The paper uses a recently developed state-of-the-art method for few-shot incremental learning of objects. After learning the object classes incrementally, the robot performs a table cleaning task organizing objects into categories specified by the human. We also demonstrate the system's ability to learn arrangements of objects and predict missing or incorrectly placed objects. Experimental evaluations demonstrate that our approach achieves nearly the same performance as a system trained with all examples at one time (batch training), which constitutes a theoretical upper bound.

Ali Ayub, Alan R. Wagner

The two main challenges faced by continual learning approaches are catastrophic forgetting and memory limitations on the storage of data. To cope with these challenges, we propose a novel, cognitively-inspired approach which trains autoencoders with Neural Style Transfer to encode and store images. Reconstructed images from encoded episodes are replayed when training the classifier model on a new task to avoid catastrophic forgetting. The loss function for the reconstructed images is weighted to reduce its effect during classifier training to cope with image degradation. When the system runs out of memory the encoded episodes are converted into centroids and covariance matrices, which are used to generate pseudo-images during classifier training, keeping classifier performance stable with less memory. Our approach increases classification accuracy by 13-17% over state-of-the-art methods on benchmark datasets, while requiring 78% less storage space.

Kasra Mokhtari, Alan R. Wagner

We present the pedestrian patterns dataset for autonomous driving. The dataset was collected by repeatedly traversing the same three routes for one week starting at different specific timeslots. The purpose of the dataset is to capture the patterns of social and pedestrian behavior along the traversed routes at different times and to eventually use this information to make predictions about the risk associated with autonomously traveling along different routes. This dataset contains the Full HD videos and GPS data for each traversal. Fast R-CNN pedestrian detection method is applied to the captured videos to count the number of pedestrians at each video frame in order to assess the density of pedestrians along a route. By providing this large-scale dataset to researchers, we hope to accelerate autonomous driving research not only to estimate the risk, both to the public and to the autonomous vehicle but also accelerate research on long-term vision-based localization of mobile robots and autonomous vehicles of the future.

Ali Ayub, Alan R. Wagner

This paper investigates the use of game theoretic representations to represent and learn how to play interactive games such as Connect Four. We combine aspects of learning by demonstration, active learning, and game theory allowing a robot to leverage its developing representation of the game to conduct question/answer sessions with a person, thus filling in gaps in its knowledge. The paper demonstrates a method for teaching a robot the win conditions of the game Connect Four and its variants using a single demonstration and a few trial examples with a question and answer session led by the robot. Our results show that the robot can learn arbitrary win conditions for the game with little prior knowledge of the win conditions and then play the game with a human utilizing the learned win conditions. Our experiments also show that some questions are more important for learning the game's win conditions. We believe that this method could be broadly applied to a variety of interactive learning scenarios.

Ali Ayub, Alan R. Wagner

This paper contributes a novel cognitively-inspired method for RGB-D indoor scene classification. High intra-class variance and low inter-class variance make indoor scene classification an extremely challenging task. To cope with this problem, we propose a clustering approach inspired by the concept learning model of the hippocampus and the neocortex, to generate clusters and centroids for different scene categories. Test images depicting different scenes are classified by using their distance to the closest centroids (concepts). Modeling of RGB-D scenes as centroids not only leads to state-of-the-art classification performance on benchmark datasets (SUN RGB-D and NYU Depth V2), but also offers a method for inspecting and interpreting the space of centroids. Inspection of the centroids generated by our approach on RGB-D datasets leads us to propose a method for merging conceptually similar categories, resulting in improved accuracy for all approaches.