The existing deep hashing methods for image retrieval typically modeling hash-coding layer in real number space. However, these methods frequently overlook the intrinsic information loss that occurs during the hash-coding process, as the hash layer performs two tasks simultaneously: spatial transformation and dimensionality reduction. Especially in multi-label image retrieval, the exponential increase in the number of combinations with the labels further amplifies the information loss in Hamming space. Consequently, the efficiency of the hash-coding is unsatisfactory. To mitigate this limitation, we introduce a novel approach termed WalshHash, which is grounded in the principles of Walsh transformation in signal processing. Unlike conventional techniques, WalshHash formulates the hash-coding layer as a filtering process based on Kolmogorov-Arnold Networks (KANs) in the Walsh domain accompanied by constraint loss functions on multiple domains. It ensures the dimensionality reduction in the Walsh domain can be effectively projected onto the real number domain with minimal information loss, because the Walsh space encapsulates the critical information components. As a result, WalshHash demonstrates superior performance in multi-label image retrieval compared to State-of-the-Art (SOTA) methods.

We aim at creating a transfer reinforcement learning framework that allows the design of learning controllers to leverage prior knowledge extracted from previously learned tasks and previous data to improve the learning performance of new tasks. Toward this goal, we formalize knowledge transfer by expressing knowledge in the value function in our problem construct, which is referred to as reinforcement learning with knowledge shaping (RL-KS). Unlike most transfer learning studies that are empirical in nature, our results include not only simulation verifications but also an analysis of algorithm convergence and solution optimality. Also different from the well-established potential-based reward shaping methods which are built on proofs of policy invariance, our RL-KS approach allows us to advance toward a new theoretical result on positive knowledge transfer. Furthermore, our contributions include two principled ways that cover a range of realization schemes to represent prior knowledge in RL-KS. We provide extensive and systematic evaluations of the proposed RL-KS method. The evaluation environments not only include classical RL benchmark problems but also include a challenging task of real-time control of a robotic lower limb with a human user in the loop.

This is the first attempt at considering human influence in the reinforcement learning control of a robotic lower limb prosthesis toward symmetrical walking in real world situations. We propose a collaborative multi-agent reinforcement learning (cMARL) solution framework for this highly complex and challenging human-prosthesis collaboration (HPC) problem. The design of an automatic controller of the robot within the HPC context is based on accessible physical features or measurements that are known to affect walking performance. Comparisons are made with the current state-of-the-art robot control designs, which are single-agent based, as well as existing MARL solution approaches tailored to the problem, including multi-agent deep deterministic policy gradient (MADDPG) and counterfactual multi-agent policy gradient (COMA). Results show that, when compared to these approaches, treating the human and robot as coupled agents and using estimated human adaption in robot control design can achieve lower stage cost, peak error, and symmetry value to ensure better human walking performance. Additionally, our approach accelerates learning of walking tasks and increases learning success rate. The proposed framework can potentially be further developed to examine how human and robotic lower limb prosthesis interact, an area that little is known about. Advancing cMARL toward real world applications such as HPC for normative walking sets a good example of how AI can positively impact on people’s lives.

We are motivated by the real challenges presented in a human-robot system to develop new designs that are efficient at data level and with performance guarantees, such as stability and optimality at system level. Existing approximate/adaptive dynamic programming (ADP) results that consider system performance theoretically are not readily providing practically useful learning control algorithms for this problem, and reinforcement learning (RL) algorithms that address the issue of data efficiency usually do not have performance guarantees for the controlled system. This study fills these important voids by introducing innovative features to the policy iteration algorithm. We introduce flexible policy iteration (FPI), which can flexibly and organically integrate experience replay and supplemental values from prior experience into the RL controller. We show system-level performances, including convergence of the approximate value function, (sub)optimality of the solution, and stability of the system. We demonstrate the effectiveness of the FPI via realistic simulations of the human-robot system. It is noted that the problem we face in this study may be difficult to address by design methods based on classical control theory as it is nearly impossible to obtain a customized mathematical model of a human-robot system either online or offline. The results we have obtained also indicate the great potential of RL control to solving realistic and challenging problems with high-dimensional control inputs.

Personalizing medical devices such as lower limb wearable robots is challenging. While the initial feasibility of automating the process of knee prosthesis control parameter tuning has been demonstrated in a principled way, the next critical issue is to improve tuning efficiency and speed it up for the human user, in clinic settings, while maintaining human safety. We, therefore, propose a policy iteration with constraint embedded (PICE) method as an innovative solution to the problem under the framework of reinforcement learning. Central to PICE is the use of a projected Bellman equation with a constraint of assuring positive semidefiniteness of performance values during policy evaluation. Additionally, we developed both online and offline PICE implementations that provide additional flexibility for the designer to fully utilize measurement data, either from on-policy or off-policy, to further improve PICE tuning efficiency. Our human subject testing showed that the PICE provided effective policies with significantly reduced tuning time. For the first time, we also experimentally evaluated and demonstrated the robustness of the deployed policies by applying them to different tasks and users. Putting it together, our new way of problem solving has been effective as PICE has demonstrated its potential toward truly automating the process of control parameter tuning for robotic knee prosthesis users.

Robotic prostheses provide new opportunities to better restore lost functions than passive prostheses for trans-femoral amputees. But controlling a prosthesis device automatically for individual users in different task environments is an unsolved problem. Reinforcement learning (RL) is a naturally promising tool. For prosthesis control with a user in the loop, it is desirable that the controlled prosthesis can adapt to different task environments as quickly and smoothly as possible. However, most RL agents learn or relearn from scratch when the environment changes. To address this issue, we propose the knowledge-guided Q-learning (KG-QL) control method as a principled way for the problem. In this report, we collected and used data from two able-bodied (AB) subjects wearing a RL controlled robotic prosthetic limb walking on level ground. Our ultimate goal is to build an efficient RL controller with reduced time and data requirements and transfer knowledge from AB subjects to amputee subjects. Toward this goal, we demonstrate its feasibility by employing OpenSim, a well-established human locomotion simulator. Our results show the OpenSim simulated amputee subject improved control tuning performance over learning from scratch by utilizing knowledge transfer from AB subjects. Also in this paper, we will explore the possibility of information transfer from AB subjects to help tuning for the amputee subjects.

Robotic prostheses deliver greater function than passive prostheses, but we face the challenge of tuning a large number of control parameters in order to personalize the device for individual amputee users. This problem is not easily solved by traditional control designs or the latest robotic technology. Reinforcement learning (RL) is naturally appealing. The recent, unprecedented success of AlphaZero demonstrated RL as a feasible, large-scale problem solver. However, the prosthesis-tuning problem is associated with several unaddressed issues such as that it does not have a known and stable model, the continuous states and controls of the problem may result in a curse of dimensionality, and the human-prosthesis system is constantly subject to measurement noise, environmental change and human-body-caused variations. In this paper, we demonstrated the feasibility of direct heuristic dynamic programming, an approximate dynamic programming (ADP) approach, to automatically tune the 12 robotic knee prosthesis parameters to meet individual human users' needs. We tested the ADP-tuner on two subjects (one able-bodied subject and one amputee subject) walking at a fixed speed on a treadmill. The ADP-tuner learned to reach target gait kinematics in an average of 300 gait cycles or 10 min of walking. We observed improved ADP tuning performance when we transferred a previously learned ADP controller to a new learning session with the same subject. To the best of our knowledge, our approach to personalize robotic prostheses is the first implementation of online ADP learning control to a clinical problem involving human subjects.

This paper aims to develop an optimal controller that can automatically provide personalized control of robotic knee prosthesis in order to best support gait of individual prosthesis wearers. We introduced a new reinforcement learning (RL) controller for this purpose based on the promising ability of RL controllers to solve optimal control problems through interactions with the environment without requiring an explicit system model. However, collecting data from a human-prosthesis system is expensive and thus the design of a RL controller has to take into account data and time efficiency. We therefore propose an offline policy iteration based reinforcement learning approach. Our solution is built on the finite state machine (FSM) impedance control framework, which is the most used prosthesis control method in commercial and prototypic robotic prosthesis. Under such a framework, we designed an approximate policy iteration algorithm to devise impedance parameter update rules for 12 prosthesis control parameters in order to meet individual users' needs. The goal of the reinforcement learning-based control was to reproduce near-normal knee kinematics during gait. We tested the RL controller obtained from offline learning in real time experiment involving the same able-bodied human subject wearing a robotic lower limb prosthesis. Our results showed that the RL control resulted in good convergent behavior in kinematic states, and the offline learning control policy successfully adjusted the prosthesis control parameters to produce near-normal knee kinematics in 10 updates of the impedance control parameters.

This brief presents a novel application of adaptive dynamic programming (ADP) for optimal adaptive control of powered lower limb prostheses, a type of wearable robots to assist the motor function of the limb amputees. Current control of these robotic devices typically relies on finite state impedance control (FS-IC), which lacks adaptability to the user's physical condition. As a result, joint impedance settings are often customized manually and heuristically in clinics, which greatly hinder the wide use of these advanced medical devices. This simulation study aimed at demonstrating the feasibility of ADP for automatic tuning of the twelve knee joint impedance parameters during a complete gait cycle to achieve balanced walking. Given that the accurate models of human walking dynamics are difficult to obtain, the model-free ADP control algorithms were considered. First, direct heuristic dynamic programming (dHDP) was applied to the control problem, and its performance was evaluated on OpenSim, an often-used dynamic walking simulator. For the comparison purposes, we selected another established ADP algorithm, the neural fitted Q with continuous action (NFQCA). In both cases, the ADP controllers learned to control the right knee joint and achieved balanced walking, but dHDP outperformed NFQCA in this application during a 200 gait cycle-based testing.