Preference-Guided Reinforcement Learning for Efficient Exploration

In this paper, we investigate preference-based reinforcement learning (PbRL) that allows reinforcement learning (RL) agents to learn from human feedback. This is particularly valuable when defining a fine-grain reward function is not feasible. However, this approach is inefficient and impractical for promoting deep exploration in hard-exploration tasks with long horizons and sparse rewards. To tackle this issue, we introduce LOPE: Learning Online with trajectory Preference guidancE, an end-to-end preference-guided RL framework that enhances exploration efficiency in hard-exploration tasks. Our intuition is that LOPE directly adjusts the focus of online exploration by considering human feedback as guidance, avoiding learning a separate reward model from preferences. Specifically, LOPE includes a two-step sequential policy optimization process consisting of trust-region-based policy improvement and preference guidance steps. We reformulate preference guidance as a novel trajectory-wise state marginal matching problem that minimizes the maximum mean discrepancy distance between the preferred trajectories and the learned policy. Furthermore, we provide a theoretical analysis to characterize the performance improvement bound and evaluate the LOPE's effectiveness. When assessed in various challenging hard-exploration environments, LOPE outperforms several state-of-the-art methods regarding convergence rate and overall performance. The code used in this study is available at \url{https://github.com/buaawgj/LOPE}.

Contrastive Learning of Shared Spatiotemporal EEG Representations Across Individuals for Naturalistic Neuroscience

Neural representations induced by naturalistic stimuli offer insights into how humans respond to peripheral stimuli in daily life. The key to understanding the general neural mechanisms underlying naturalistic stimuli processing involves aligning neural activities across individuals and extracting inter-subject shared neural representations. Targeting the Electroencephalogram (EEG) technique, known for its rich spatial and temporal information, this study presents a general framework for Contrastive Learning of Shared SpatioTemporal EEG Representations across individuals (CL-SSTER). Harnessing the representational capabilities of contrastive learning, CL-SSTER utilizes a neural network to maximize the similarity of EEG representations across individuals for identical stimuli, contrasting with those for varied stimuli. The network employed spatial and temporal convolutions to simultaneously learn the spatial and temporal patterns inherent in EEG. The versatility of CL-SSTER was demonstrated on three EEG datasets, including a synthetic dataset, a speech audio EEG dataset, and an emotional video EEG dataset. CL-SSTER attained the highest inter-subject correlation (ISC) values compared to the state-of-the-art ISC methods. The latent representations generated by CL-SSTER exhibited reliable spatiotemporal EEG patterns, which can be explained by specific aspects of the stimuli. CL-SSTER serves as an interpretable and scalable foundational framework for the identification of inter-subject shared neural representations in the realm of naturalistic neuroscience.

Optimization Landscape of Policy Gradient Methods for Discrete-time Static Output Feedback

In recent times, significant advancements have been made in delving into the optimization landscape of policy gradient methods for achieving optimal control in linear time-invariant (LTI) systems. Compared with state-feedback control, output-feedback control is more prevalent since the underlying state of the system may not be fully observed in many practical settings. This paper analyzes the optimization landscape inherent to policy gradient methods when applied to static output feedback (SOF) control in discrete-time LTI systems subject to quadratic cost. We begin by establishing crucial properties of the SOF cost, encompassing coercivity, L-smoothness, and M-Lipschitz continuous Hessian. Despite the absence of convexity, we leverage these properties to derive novel findings regarding convergence (and nearly dimension-free rate) to stationary points for three policy gradient methods, including the vanilla policy gradient method, the natural policy gradient method, and the Gauss-Newton method. Moreover, we provide proof that the vanilla policy gradient method exhibits linear convergence towards local minima when initialized near such minima. The paper concludes by presenting numerical examples that validate our theoretical findings. These results not only characterize the performance of gradient descent for optimizing the SOF problem but also provide insights into the effectiveness of general policy gradient methods within the realm of reinforcement learning.

Box2Poly: Memory-Efficient Polygon Prediction of Arbitrarily Shaped and Rotated Text

Recently, Transformer-based text detection techniques have sought to predict polygons by encoding the coordinates of individual boundary vertices using distinct query features. However, this approach incurs a significant memory overhead and struggles to effectively capture the intricate relationships between vertices belonging to the same instance. Consequently, irregular text layouts often lead to the prediction of outlined vertices, diminishing the quality of results. To address these challenges, we present an innovative approach rooted in Sparse R-CNN: a cascade decoding pipeline for polygon prediction. Our method ensures precision by iteratively refining polygon predictions, considering both the scale and location of preceding results. Leveraging this stabilized regression pipeline, even employing just a single feature vector to guide polygon instance regression yields promising detection results. Simultaneously, the leverage of instance-level feature proposal substantially enhances memory efficiency (>50% less vs. the state-of-the-art method DPText-DETR) and reduces inference speed (>40% less vs. DPText-DETR) with minor performance drop on benchmarks.

Trust-Region Neural Moving Horizon Estimation for Robots

Accurate disturbance estimation is essential for safe robot operations. The recently proposed neural moving horizon estimation (NeuroMHE), which uses a portable neural network to model the MHE's weightings, has shown promise in further pushing the accuracy and efficiency boundary. Currently, NeuroMHE is trained through gradient descent, with its gradient computed recursively using a Kalman filter. This paper proposes a trust-region policy optimization method for training NeuroMHE. We achieve this by providing the second-order derivatives of MHE, referred to as the MHE Hessian. Remarkably, we show that much of computation already used to obtain the gradient, especially the Kalman filter, can be efficiently reused to compute the MHE Hessian. This offers linear computational complexity relative to the MHE horizon. As a case study, we evaluate the proposed trust region NeuroMHE on real quadrotor flight data for disturbance estimation. Our approach demonstrates highly efficient training in under 5 min using only 100 data points. It outperforms a state-of-the-art neural estimator by up to 68.1% in force estimation accuracy, utilizing only 1.4% of its network parameters. Furthermore, our method showcases enhanced robustness to network initialization compared to the gradient descent counterpart.

Efficient Q-Learning over Visit Frequency Maps for Multi-agent Exploration of Unknown Environments

The robot exploration task has been widely studied with applications spanning from novel environment mapping to item delivery. For some time-critical tasks, such as rescue catastrophes, the agent is required to explore as efficiently as possible. Recently, Visit Frequency-based map representation achieved great success in such scenarios by discouraging repetitive visits with a frequency-based penalty. However, its relatively large size and single-agent settings hinder its further development. In this context, we propose Integrated Visit Frequency Map, which encodes identical information as Visit Frequency Map with a more compact size, and a visit frequency-based multi-agent information exchange and control scheme that is able to accommodate both representations. Through tests in diverse settings, the results indicate our proposed methods can achieve a comparable level of performance of VFM with lower bandwidth requirements and generalize well to different multi-agent setups including real-world environments.

Finite-time analysis of single-timescale actor-critic

Despite the great empirical success of actor-critic methods, its finite-time convergence is still poorly understood in its most practical form. In particular, the analysis of single-timescale actor-critic presents significant challenges due to the highly inaccurate critic estimation and the complex error propagation dynamics over iterations. Existing works on analyzing single-timescale actor-critic only focus on the i.i.d. sampling or tabular setting for simplicity, which is rarely the case in practical applications. We consider the more practical online single-timescale actor-critic algorithm on continuous state space, where the critic is updated with a single Markovian sample per actor step. We prove that the online single-timescale actor-critic method is guaranteed to find an $\epsilon$-approximate stationary point with $\widetilde{\mathcal{O}}(\epsilon^{-2})$ sample complexity under standard assumptions, which can be further improved to $\mathcal{O}(\epsilon^{-2})$ under i.i.d. sampling. Our analysis develops a novel framework that evaluates and controls the error propagation between actor and critic in a systematic way. To our knowledge, this is the first finite-time analysis for online single-timescale actor-critic method. Overall, our results compare favorably to the existing literature on analyzing actor-critic in terms of considering the most practical settings and requiring weaker assumptions.

Global Convergence of Two-timescale Actor-Critic for Solving Linear Quadratic Regulator

The actor-critic (AC) reinforcement learning algorithms have been the powerhouse behind many challenging applications. Nevertheless, its convergence is fragile in general. To study its instability, existing works mostly consider the uncommon double-loop variant or basic models with finite state and action space. We investigate the more practical single-sample two-timescale AC for solving the canonical linear quadratic regulator (LQR) problem, where the actor and the critic update only once with a single sample in each iteration on an unbounded continuous state and action space. Existing analysis cannot conclude the convergence for such a challenging case. We develop a new analysis framework that allows establishing the global convergence to an $\epsilon$-optimal solution with at most an $\tilde{\mathcal{O}}(\epsilon^{-2.5})$ sample complexity. To our knowledge, this is the first finite-time convergence analysis for the single sample two-timescale AC for solving LQR with global optimality. The sample complexity improves those of other variants by orders, which sheds light on the practical wisdom of single sample algorithms. We also further validate our theoretical findings via comprehensive simulation comparisons.

AXNet: ApproXimate computing using an end-to-end trainable neural network

Neural network based approximate computing is a universal architecture promising to gain tremendous energy-efficiency for many error resilient applications. To guarantee the approximation quality, existing works deploy two neural networks (NNs), e.g., an approximator and a predictor. The approximator provides the approximate results, while the predictor predicts whether the input data is safe to approximate with the given quality requirement. However, it is non-trivial and time-consuming to make these two neural network coordinate---they have different optimization objectives---by training them separately. This paper proposes a novel neural network structure---AXNet---to fuse two NNs to a holistic end-to-end trainable NN. Leveraging the philosophy of multi-task learning, AXNet can tremendously improve the invocation (proportion of safe-to-approximate samples) and reduce the approximation error. The training effort also decrease significantly. Experiment results show 50.7% more invocation and substantial cuts of training time when compared to existing neural network based approximate computing framework.