Recognising intent in collaborative human robot tasks can improve team performance and human perception of robots. Intent can differ from the observed outcome in the presence of mistakes which are likely in physically dynamic tasks. We created a dataset of 1227 throws of a ball at a target from 10 participants and observed that 47% of throws were mistakes with 16% completely missing the target. Our research leverages facial images capturing the person's reaction to the outcome of a throw to predict when the resulting throw is a mistake and then we determine the actual intent of the throw. The approach we propose for outcome prediction performs 38% better than the two-stream architecture used previously for this task on front-on videos. In addition, we propose a 1-D CNN model which is used in conjunction with priors learned from the frequency of mistakes to provide an end-to-end pipeline for outcome and intent recognition in this throwing task.

This paper investigates how risk influences the way people barter. We used Minecraft to create an experimental environment in which people bartered to earn a monetary bonus. Our findings reveal that subjects exhibit risk-aversion to competitive bartering environments and deliberate over their trades longer when compared to cooperative environments. These initial experiments lay groundwork for development of agents capable of strategically trading with human counterparts in different environments.

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.

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 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. A primary challenge faced by robots deployed in the real world is continual adaptation to dynamic environments. Central to this challenge is object recognition (Ayub & Wagner, 2020d), a task typically requiring labeled examples. In this work, we address the problem of parsimonious object labelling wherein a robot may request labels for a small number of objects about which it knows least. In recent years, several works have been directed toward the problem of Few-Shot Class Incremental Learning (FSCIL) (Tao et al., 2020; Ayub & Wagner, 2020c) to develop models of incremental object learning that can learn from limited training data for each object class. The literature has made significant progress toward developing robots that can continually learn new objects from limited training data while preserving knowledge of previous objects. However, existing methods make strong assumptions about the training data that are rarely true in the real world. For example, FSCIL assumes that in each increment the robot will receive a fully labeled image dataset for the object classes in that increment, and the robot will not receive more data for these classes again (Tao et al., 2020; Ayub & Wagner, 2020c;d). In real world environments, however, robots will most likely encounter many unlabeled objects in their environment, and they will have to direct their learning toward a smaller subset of those unknown objects. Active learning is a subfield of machine learning that focuses on improving the learning efficiency of models by selectively seeking labels from within a large unlabeled data pool (Settles, 2009; Ayub & Fendley, 2022). Related to active learning is active class selection (ACS) in which a model seeks labels for specific object classes (Lomasky et al., 2007).

Recent research has examined the possibility of using robots to guide evacuees to safe exits during emergencies. Yet, there are many factors that can impact a person's decision to follow a robot. Being able to model how an evacuee follows an emergency robot guide could be crucial for designing robots that effectively guide evacuees during an emergency. This paper presents a method for developing realistic and predictive human evacuee models from physical human evacuation experiments. The paper analyzes the behavior of 14 human subjects during physical robot-guided evacuation. We then use the video data to create evacuee motion models that predict the person's future positions during the emergency. Finally, we validate the resulting models by running a k-fold cross-validation on the data collected during physical human subject experiments. We also present performance results of the model using data from a similar simulated emergency evacuation experiment demonstrating that these models can serve as a tool to predict evacuee behavior in novel evacuation simulations.

Emergency evacuation describes a complex situation involving time-critical decision-making by evacuees. Mobile robots are being actively explored as a potential solution to provide timely guidance. In this work, we study a robot-guided crowd evacuation problem where a small group of robots is used to guide a large human crowd to safe locations. The challenge lies in how to utilize micro-level human-robot interactions to indirectly influence a population that significantly outnumbers the robots to achieve the collective evacuation objective. To address the challenge, we follow a two-scale modeling strategy and explore mean-field hydrodynamic models which consist of a family of microscopic social-force models that explicitly describe how human movements are locally affected by other humans, the environment, and the robots, and associated macroscopic equations for the temporal and spatial evolution of the crowd density and flow velocity. We design controllers for the robots such that they not only automatically explore the environment (with unknown dynamic obstacles) to cover it as much as possible but also dynamically adjust the directions of their local navigation force fields based on the real-time macro-states of the crowd to guide the crowd to a safe location. We prove the stability of the proposed evacuation algorithm and conduct a series of simulations (involving unknown dynamic obstacles) to validate the performance of the algorithm.

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.

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.

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. Humans continue to learn new concepts over their lifetime without the need to relearn most previous concepts. Modern machine learning systems, however, require the complete training data to be available at one time (batch learning) (Girshick, 2015). In this paper, we consider the problem of continual learning from the class-incremental perspective. Class-incremental systems are required to learn from a stream of data belonging to different classes and are evaluated in a single-headed evaluation (Chaudhry et al., 2018). In single-headed evaluation, the model is evaluated on all classes observed so far without any information indicating which class is being observed. Creating highly accurate class-incremental learning systems is a challenging problem.

Classifying images taken from indoor scenes is an important area of research. The development of an accurate indoor scene classifier has the potential to improve indoor localization and decision-making for domestic robots, offer new applications for wearable computer users, and generally result in better vision-based situation awareness thus impacting a wide variety of applications. The introduction of deep learning methods, the creation of numerous large-scale datasets, and the development of specialized computing hardware have all contributed to the rapid improvement in image classification performance. One reason for deep learning's success has been the ability to learn multiple layers of generic image features that can then be used on other related computer vision problems. For instance, features from object trained image classifiers have been used to train indoor scene classifiers [27]. Yet, indoor scene classification is a challenging problem on its own.