Unsupervised domain adaptation (UDA) aims to transfer knowledge from a labeled source domain to an unlabeled target domain. Recent works have focused on source-free UDA, where only target data is available. This is challenging as models rely on noisy pseudo-labels and struggle with distribution shifts. We propose Active Adversarial Alignment (A3), a novel framework combining self-supervised learning, adversarial training, and active learning for robust source-free UDA. A3 actively samples informative and diverse data using an acquisition function for training. It adapts models via adversarial losses and consistency regularization, aligning distributions without source data access. A3 advances source-free UDA through its synergistic integration of active and adversarial learning for effective domain alignment and noise reduction. Our approach significantly advances state-of-the-art methods, achieving 4.1% on Office-31, 11.7% on Office-Home, and 10.6% on DomainNet accuracy improvements. Source code: https://github.com/chrisantuseze/active-self-pretraining

This paper explores automated modification of text to make it more accessible for people with dyslexia, a reading disorder affecting a significant percentage of the global population. The modifications are both in terms of changing the appearance of text and simplification of the words, grammar, and length of textual material. For simplification of text, we built a dataset with original and dyslexia-friendly text verified by human readers that it improve their reading experience by 27% on average. Then we developed a pipeline to generate dyslexia-friendly text automatically using transfer learning. The model learns styles appropriate for dyslexic users and generates dyslexia-friendly text from arbitrary textual data, which is easier for people with dyslexia to read and interpret.

There have been many advances in the artificial intelligence field due to the emergence of deep learning. In almost all sub-fields, artificial neural networks have reached or exceeded human-level performance. However, most of the models are not interpretable. As a result, it is hard to trust their decisions, especially in life and death scenarios. In recent years, there has been a movement toward creating explainable artificial intelligence, but most work to date has concentrated on image processing models, as it is easier for humans to perceive visual patterns. There has been little work in other fields like natural language processing. In this paper, we train a convolutional model on textual data and analyze the global logic of the model by studying its filter values. In the end, we find the most important words in our corpus to our models logic and remove the rest (95%). New models trained on just the 5% most important words can achieve the same performance as the original model while reducing training time by more than half. Approaches such as this will help us to understand NLP models, explain their decisions according to their word choices, and improve them by finding blind spots and biases.

3D printed prosthetics have narrowed the gap between the tens of thousands of dollars cost of traditional prosthetic designs and amputees’ needs. However, the World Health Organization estimates that only 5-15% of people can receive adequate prosthesis services [2]. To resolve the lack of prosthesis supply and reduce cost issues (for both materials and maintenance), this paper provides an overview of a self-trainable user-customized system architecture for a 3D printed prosthetic hand to minimize the challenge of accessing and maintaining these supporting devices. In this paper, we develop and implement a customized behavior system that can generate any gesture that users desire. The architecture provides upper limb amputees with self-trainable software and can improve their prosthetic performance at almost no financial cost. All kinds of unique gestures that users want are trainable with the RBF network using 3 channel EMG sensor signals with a 94% average success rate. This result demonstrates that applying user-customized training to the behavior of a prosthetic hand can satisfy individual user requirements in real-life activities with high performance.

In this article, we demonstrate the ability to recognize hand gestures in a noncontact wireless fashion using only incoherent light signals reflected from a human subject. Fundamentally distinguished from radar, lidar, and camera-based sensing systems, this sensing modality uses only a low-cost light source (e.g., LED) and a sensor (e.g., photodetector). The lightwave-based gesture recognition system identifies different gestures from the variations in light intensity reflected from the subject's hand within a short (20-35 cm) range. As users perform different gestures, scattered light forms unique, statistically repeatable, time-domain signatures. These signatures can be learned by repeated sampling to obtain the training model against which unknown gesture signals are tested and categorized. These time-domain variations of the lightwave signals reflected from hand are denoised, standardized, and then classified by using machine learning classification tools such as $K$-nearest neighbors and support vector machine. Performance evaluations have been conducted with eight gestures, five subjects, different distances and lighting conditions, and visible and infrared light sources. The results demonstrate the best hand gesture recognition performance of infrared sensing at 20 cm with an average of 96% accuracy. The developed gesture recognition system is low-cost, effective, and noncontact technology for numerous human-computer interaction applications.

The researchers in this study have developed a novel approach using mutual reinforcement learning (MRL) where both the robot and human act as empathetic individuals who function as reinforcement learning agents for each other to achieve a particular task over continuous communication and feedback. This shared model not only has a collective impact but improves human cognition and helps in building a successful human-robot relationship. In our current work, we compared our learned reinforcement model with a baseline non-reinforcement and random approach in a robotics domain to identify the significance and impact of MRL. MRL contributed to improved skill transfer, and the robot was able successfully to predict which reinforcement behaviors would be most valuable to its human partners.

Robots are capable of training humans to achieve complex tasks, and their helpful feedback can lead to useful human-robot collaborations. In this research we present a reinforcement learning model influenced by human cognition which is repurposed to enhance human learning, investigate a robot's ability to encourage and motivate humans and improve their performance. During teaching the robot trades off between exploration and exploitation to understand the human perception and develop a successful motivational approach. We compare our learned reinforcement model with a baseline nonreinforcement approach and with a random reinforcer, and achieve more effective teaching in the learned reinforcement condition. In addition, we discovered an extremely strong relationship (r = 0.88) between the robot's regret, in a machine learning sense, and the performance of its human partner.

Instructing human novices on complex tasks in non-standardized environments are an underexplored potential use for social co-robots, since instruction and skill transfer involving human experts can require an enormous commitment of time and resources. In this paper, we enable a humanoid Baxter robot to build a semantically accessible framework for task learning, teaching and representation via active learning with human experts using hierarchical semantic labels. This process not only helps the robot to learn tasks from expert demonstrations, but later improves the ability of the robot to teach novice human operators. Our results show that the better-understood learning from demonstration (LfD) task is greatly enhanced by the active learning and mutual semantic structure building in a expert-robot partnership, while the robot's ability to teach novices is improved, though the results are suggestive rather than conclusive at this point. We discuss the important aspects and power of learning and teaching from demonstration and how both benefit from communication and joint human-robot creation of semantic hierarchies.

A growing concern with deploying Unmanned Aerial Vehicles (UAVs) in urban environments is the potential violation of human privacy, and the backlash this could entail. Therefore, there is a need for UAV path planning algorithms that minimize the likelihood of invading human privacy. We formulate the problem of human-aware path planning as a nonstationary Markov Decision Process, and provide a novel model-based reinforcement learning solution that leverages Gaussian process clustering. Our algorithm is flexible enough to accommodate changes in human population densities by employing Bayesian nonparametrics, and is real-time computable. The approach is validated experimentally on a large-scale long duration experiment with both simulated and real UAVs.

The advent of ROS, the Robot Operating System, has finally made it possible to implement and use state-of-the-art navigation and manipulation algorithms on widely-available, inexpensive standard robot platforms. With the addition of the Rosbridge application programming interface, interface designers and applications programmers can create robot interfaces and behaviors without venturing into the specialized world of robotics engineers. This tutorial introduces ROS and Rosbridge, and shows how quickly and easily these tools can be used to design and conduct large-scale online HRI experiments, access algorithms for autonomous robot behavior, and leverage the huge ecosystem of general-purpose web-based and application-oriented software engineering for robotics and HRI research. Tutorial attendees will learn the basics of autonomous and teleoperated navigation and manipulation, as well as interface design for online interaction with robots. During the tutorial they will design and write their own remote presence application, as well as develop strategies for incorporating autonomy and dealing with data collection.

In this paper, we describe a remote lab system that allows remote groups to access a shared PR2. This lab will enable a larger and more diverse group of researchers to participate directly in state-of-the-art robotics research and will improve the reproducibility and comparability of robotics experiments. We identify a set of requirements that apply to all web-based remote laboratories and focus on solutions to these requirements. Specifically, we present solutions to interface, control and design difficulties in the client and server-side software when implementing a remote laboratory architecture. The combination of shared physical hardware and shared middleware software allows for experiments that build upon and compare against results on the same platform and in the same environment for common tasks. We describe how researchers can interact with the PR2 and its environment remotely through a web interface, as well as develop similar interfaces to visualize and run experiments remotely.

We present a study of using a robotic learning from demonstration system capable of collecting large amounts of human-robot interaction data through a web-based interface. We examine the effect of different perceptual mappings between the human teacher and robot on the learning from demonstration. We show that humans are significantly more effective at teaching a robot to navigate a maze when presented with information that is limited to the robot's perception of the world, even though their task performance measurably suffers when contrasted with users provided with a natural and detailed raw video feed. Robots trained on such demonstrations learn more quickly, perform more accurately and generalize better. We also demonstrate a set of software tools for enabling internet-mediated human-robot interaction and gathering the large datasets that such crowdsourcing makes possible.

We describe our efforts to create infrastructure to enable web interfaces for robotics. Such interfaces will enable researchers and users to remotely access robots through the internet as well as expand the types of robotic applications available to users with web-enabled devices. This paper centers on rosjs, a lightweight Javascript binding for ROS, Willow Garage's robot middleware framework, rosjs exposes many of the capabilities of ROS, allowing application developers to write controllers that are executed through a web browser. We discuss how rosjs extends ROS and briefly overview some of the features it provides, rosjs has been instrumental in the creation of remote laboratories featuring the iRobot Create and the PR2. These facilities will be available to the community as experimental resources. We describe the overall goals of this project as well as provide a brief description of how rosjs was used to help create web interfaces for these facilities.

Bayesian networks (BNs) have been widely used as a model for knowledge representation and probabilistic inferences. However, the single probability representation of conditional dependencies has been proven to be over-constrained in realistic applications. Many efforts have proposed to represent the dependencies using probability intervals instead of single probabilities. In this paper, we move one step further and adopt a probability distribution schema. This results in a higher order representation of uncertainties in a BN. We formulate probabilistic inferences in this context and then propose a mean/covariance propagation algorithm based on the well-known junction tree propagation for standard BNs. For algorithm validation, we develop a two-layered Markov likelihood weighting approach that handles high-order uncertainties and provides "ground-truth" solutions to inferences, albeit very slowly. Our experiments show that the mean/covariance propagation algorithm can efficiently produce high-quality solutions that compare favorably to results obtained through painstaking sampling

Music performance is an important, well-structured setting for evaluating a robot's ability to detect, understand and respond appropriately to complex human activity. Social tasks such as cooperative performance require participants to detect, interpret and attune to the actions of their partners quickly and accurately. The synthesis of multiple sensory perceptions may be a fruitful approach to this problem. In order to evaluate this approach, we programmed a humanoid robot, Nico, to play a drum in concert with human drummers and at the direction of a human conductor. Our results show that sensory integration can enable precise synchronization in social tasks even when perceptual data is imperfect, misleading and subject to extensive processing delay. By integrating several streams of information - visual, auditory, and proprioceptive - Nico can attune to a tempo that is set by a human conductor, in concert with human performers. Nico continuously evaluates its perceptions of its own actions and those of the humans around it, dealing with unforeseen changes in tempo and affect in real time