Abstract
This research addresses the trainability of Parameterized Quantum Circuit-based Softmax policies in Reinforcement Learning. We assess the trainability of these policies by examining the scaling of the expected value of the partial derivative of the log policy objective function. Here, we assume the hardware-efficient ansatz with blocks forming local 2-designs. In this setting, we show that if each expectation value representing the action's numerical preference is composed of a global observable, it leads to exponentially vanishing gradients. In contrast, for n-qubit systems, if the observables are log(n)-local, the gradients vanish polynomially with the number of qubits provided O(log n) depth. We also show that the expectation of the gradient of the log policy objective depend on the entire action space. Thus, even though global observables lead to concentration, the gradient signal can still be propagated in the presence of at least a single local observable. We validate the theoretical predictions in a series of ansatze and evaluate the performance of local and global observables in a multi-armed bandit setting.

Abstract

Abstract
Artificial I ntelligence ( AI) a nd M achine Learning are frequently used to develop player skills in robotic soccer scenarios. Despite the potential of deep reinforcement learning, its computational demands pose challenges when learning complex behaviors. This work explores less demanding methods, namely Evolution Strategies (ES) and Hierarchical Reinforcement Learning (HRL), for enhancing coordination and cooperation between two agents from the FC Portugal 3D Simulation Soccer Team, in RoboCup. The goal is for two robots to learn a high-level skill that enables a robot to pass the ball to its teammate as quickly as possible. Results show that the trained models under-performed in a traditional robotic soccer two-agent task and scored perfectly in a much simpler one. Therefore, this work highlights that while these alternative methods can learn trivial cooperative behavior, more complex tasks are difficult t o learn.

Abstract
RGB-D sensors face multiple challenges operating under open-field environments because of their sensitivity to external perturbations such as radiation or rain. Multiple works are approaching the challenge of perceiving the three-dimensional (3D) position of objects using monocular cameras. However, most of these works focus mainly on deep learning-based solutions, which are complex, data-driven, and difficult to predict. So, we aim to approach the problem of predicting the three-dimensional (3D) objects’ position using a Gaussian viewpoint estimator named best viewpoint estimator (BVE), powered by an extended Kalman filter (EKF). The algorithm proved efficient on the tasks and reached a maximum average Euclidean error of about 32mm. The experiments were deployed and evaluated in MATLAB using artificial Gaussian noise. Future work aims to implement the system in a robotic system.

Abstract
Sensing at the single cell level can provide insights into its dynamics and heterogeneity, yielding information otherwise unattainable with traditional biological methods where average population behavior is observed. In this context, optical tweezers provide the ability to select, separate, manipulate and identify single cells or other types of microparticles, potentially enabling single cell diagnostics. Forward or backscatter analysis of the light interacting with the trapped cells can provide valuable insights on the cell optical, geometrical and mechanical properties. In particular, the combination of tweezers systems with advanced machine learning algorithms can enable single cell identification capabilities. However, typical processing pipelines require a training stage which often struggles when trying to generalize to new sets of data. In this context, fully automated tweezers system can provide mechanisms to obtain much larger datasets with minimum effort form the users, while eliminating procedural variability. In this work, a pipeline for full automation of optical tweezers systems is discussed. A performance comparison between manually operated and fully automated tweezers systems is presented, clearly showing advantages of the latter. A case study demonstrating the ability of the system to discriminate molecular binding events on microparticles is presented.

Abstract
Microtasks have become increasingly popular in the digital labor market since they provide easy access to a crowd of people with varying skills and aptitudes to perform remote work tasks that even the most capable algorithmic systems are unable to complete in a timely and efficient fashion. However, despite the latest advancements in crowd-powered and contiguous interfaces, many crowd workers still face some accessibility issues, which ultimately deteriorate the quality of the work produced. To mitigate this problem, we restrict attention to the development of two different web-based mini-games with a focus on cognitive personalization. We have conducted a pilot gamified experience, with six participants with autism, dyslexia, and attention deficit hyperactivity. The results suggest that a web-based mini-game can be incorporated in preliminary microtask-based crowdsourcing execution stages to achieve enhanced cognitive personalization in crowdsourcing settings.