Learning to Control Camera Exposure via Reinforcement Learning

Adjusting camera exposure in arbitrary lighting conditions is the first step to ensure the functionality of computer vision applications. Poorly adjusted camera exposure often leads to critical failure and performance degradation. Traditional camera exposure control methods require multiple convergence steps and time-consuming processes, making them unsuitable for dynamic lighting conditions. In this paper, we propose a new camera exposure control framework that rapidly controls camera exposure while performing real-time processing by exploiting deep reinforcement learning. The proposed framework consists of four contributions: 1) a simplified training ground to simulate real-world's diverse and dynamic lighting changes, 2) flickering and image attribute-aware reward design, along with lightweight state design for real-time processing, 3) a static-to-dynamic lighting curriculum to gradually improve the agent's exposure-adjusting capability, and 4) domain randomization techniques to alleviate the limitation of the training ground and achieve seamless generalization in the wild.As a result, our proposed method rapidly reaches a desired exposure level within five steps with real-time processing (1 ms). Also, the acquired images are well-exposed and show superiority in various computer vision tasks, such as feature extraction and object detection.

On Exploring PDE Modeling for Point Cloud Video Representation Learning

Point cloud video representation learning is challenging due to complex structures and unordered spatial arrangement. Traditional methods struggle with frame-to-frame correlations and point-wise correspondence tracking. Recently, partial differential equations (PDE) have provided a new perspective in uniformly solving spatial-temporal data information within certain constraints. While tracking tangible point correspondence remains challenging, we propose to formalize point cloud video representation learning as a PDE-solving problem. Inspired by fluid analysis, where PDEs are used to solve the deformation of spatial shape over time, we employ PDE to solve the variations of spatial points affected by temporal information. By modeling spatial-temporal correlations, we aim to regularize spatial variations with temporal features, thereby enhancing representation learning in point cloud videos. We introduce Motion PointNet composed of a PointNet-like encoder and a PDE-solving module. Initially, we construct a lightweight yet effective encoder to model an initial state of the spatial variations. Subsequently, we develop our PDE-solving module in a parameterized latent space, tailored to address the spatio-temporal correlations inherent in point cloud video. The process of solving PDE is guided and refined by a contrastive learning structure, which is pivotal in reshaping the feature distribution, thereby optimizing the feature representation within point cloud video data. Remarkably, our Motion PointNet achieves an impressive accuracy of 97.52% on the MSRAction-3D dataset, surpassing the current state-of-the-art in all aspects while consuming minimal resources (only 0.72M parameters and 0.82G FLOPs).

An evaluation of CFEAR Radar Odometry

This article describes the method CFEAR Radar odometry, submitted to a competition at the Radar in Robotics workshop, ICRA 20241. CFEAR is an efficient and accurate method for spinning 2D radar odometry that generalizes well across environments. This article presents an overview of the odometry pipeline with new experiments on the public Boreas dataset. We show that a real-time capable configuration of CFEAR - with its original parameter set - yields surprisingly low drift in the Boreas dataset. Additionally, we discuss an improved implementation and solving strategy that enables the most accurate configuration to run in real-time with improved robustness, reaching as low as 0.61% translation drift at a frame rate of 68 Hz. A recent release of the source code is available to the community https://github.com/dan11003/CFEAR_Radarodometry_code_public, and we publish the evaluation from this article on https://github.com/dan11003/cfear_2024_workshop

Incremental Learning with Concept Drift Detection and Prototype-based Embeddings for Graph Stream Classification

Data stream mining aims at extracting meaningful knowledge from continually evolving data streams, addressing the challenges posed by nonstationary environments, particularly, concept drift which refers to a change in the underlying data distribution over time. Graph structures offer a powerful modelling tool to represent complex systems, such as, critical infrastructure systems and social networks. Learning from graph streams becomes a necessity to understand the dynamics of graph structures and to facilitate informed decision-making. This work introduces a novel method for graph stream classification which operates under the general setting where a data generating process produces graphs with varying nodes and edges over time. The method uses incremental learning for continual model adaptation, selecting representative graphs (prototypes) for each class, and creating graph embeddings. Additionally, it incorporates a loss-based concept drift detection mechanism to recalculate graph prototypes when drift is detected.

Regularized Best-of-N Sampling to Mitigate Reward Hacking for Language Model Alignment

Best-of-N (BoN) sampling with a reward model has been shown to be an effective strategy for aligning Large Language Models (LLMs) to human preferences at the time of decoding. BoN sampling is susceptible to a problem known as reward hacking. Because the reward model is an imperfect proxy for the true objective, over-optimizing its value can compromise its performance on the true objective. A common solution to prevent reward hacking in preference learning techniques is to optimize a reward using proximity regularization (e.g., KL regularization), which ensures that the language model remains close to the reference model. In this research, we propose Regularized Best-of-N (RBoN), a variant of BoN that aims to mitigate reward hacking by incorporating a proximity term in response selection, similar to preference learning techniques. We evaluate two variants of RBoN on the AlpacaFarm dataset and find that they outperform BoN, especially when the proxy reward model has a low correlation with the true objective.

Exploring Automated Distractor Generation for Math Multiple-choice Questions via Large Language Models

Multiple-choice questions (MCQs) are ubiquitous in almost all levels of education since they are easy to administer, grade, and are a reliable format in assessments and practices. One of the most important aspects of MCQs is the distractors, i.e., incorrect options that are designed to target common errors or misconceptions among real students. To date, the task of crafting high-quality distractors largely remains a labor and time-intensive process for teachers and learning content designers, which has limited scalability. In this work, we study the task of automated distractor generation in the domain of math MCQs and explore a wide variety of large language model (LLM)-based approaches, from in-context learning to fine-tuning. We conduct extensive experiments using a real-world math MCQ dataset and find that although LLMs can generate some mathematically valid distractors, they are less adept at anticipating common errors or misconceptions among real students.

Longitudinal Targeted Minimum Loss-based Estimation with Temporal-Difference Heterogeneous Transformer

We propose Deep Longitudinal Targeted Minimum Loss-based Estimation (Deep LTMLE), a novel approach to estimate the counterfactual mean of outcome under dynamic treatment policies in longitudinal problem settings. Our approach utilizes a transformer architecture with heterogeneous type embedding trained using temporal-difference learning. After obtaining an initial estimate using the transformer, following the targeted minimum loss-based likelihood estimation (TMLE) framework, we statistically corrected for the bias commonly associated with machine learning algorithms. Furthermore, our method also facilitates statistical inference by enabling the provision of 95% confidence intervals grounded in asymptotic statistical theory. Simulation results demonstrate our method's superior performance over existing approaches, particularly in complex, long time-horizon scenarios. It remains effective in small-sample, short-duration contexts, matching the performance of asymptotically efficient estimators. To demonstrate our method in practice, we applied our method to estimate counterfactual mean outcomes for standard versus intensive blood pressure management strategies in a real-world cardiovascular epidemiology cohort study.

Compositional Estimation of Lipschitz Constants for Deep Neural Networks

The Lipschitz constant plays a crucial role in certifying the robustness of neural networks to input perturbations and adversarial attacks, as well as the stability and safety of systems with neural network controllers. Therefore, estimation of tight bounds on the Lipschitz constant of neural networks is a well-studied topic. However, typical approaches involve solving a large matrix verification problem, the computational cost of which grows significantly for deeper networks. In this letter, we provide a compositional approach to estimate Lipschitz constants for deep feedforward neural networks by obtaining an exact decomposition of the large matrix verification problem into smaller sub-problems. We further obtain a closed-form solution that applies to most common neural network activation functions, which will enable rapid robustness and stability certificates for neural networks deployed in online control settings. Finally, we demonstrate through numerical experiments that our approach provides a steep reduction in computation time while yielding Lipschitz bounds that are very close to those achieved by state-of-the-art approaches.

Multi-modal perception for soft robotic interactions using generative models

Perception is essential for the active interaction of physical agents with the external environment. The integration of multiple sensory modalities, such as touch and vision, enhances this perceptual process, creating a more comprehensive and robust understanding of the world. Such fusion is particularly useful for highly deformable bodies such as soft robots. Developing a compact, yet comprehensive state representation from multi-sensory inputs can pave the way for the development of complex control strategies. This paper introduces a perception model that harmonizes data from diverse modalities to build a holistic state representation and assimilate essential information. The model relies on the causality between sensory input and robotic actions, employing a generative model to efficiently compress fused information and predict the next observation. We present, for the first time, a study on how touch can be predicted from vision and proprioception on soft robots, the importance of the cross-modal generation and why this is essential for soft robotic interactions in unstructured environments.

An Optimization Framework to Personalize Passive Cardiac Mechanics

Personalized cardiac mechanics modeling is a powerful tool for understanding the biomechanics of cardiac function in health and disease and assisting in treatment planning. However, current models are limited to using medical images acquired at a single cardiac phase, often limiting their applicability for processing dynamic image acquisitions. This study introduces an inverse finite element analysis (iFEA) framework to estimate the passive mechanical properties of cardiac tissue using time-dependent medical image data. The iFEA framework relies on a novel nested optimization scheme, in which the outer iterations utilize a traditional optimization method to best approximate material parameters that fit image data, while the inner iterations employ an augmented Sellier's algorithm to estimate the stress-free reference configuration. With a focus on characterizing the passive mechanical behavior, the framework employs structurally based anisotropic hyperelastic constitutive models and physiologically relevant boundary conditions to simulate myocardial mechanics. We use a stabilized variational multiscale formulation for solving the governing nonlinear elastodynamics equations, verified for cardiac mechanics applications. The framework is tested in myocardium models of biventricle and left atrium derived from cardiac phase-resolved computed tomographic (CT) images of a healthy subject and three patients with hypertrophic obstructive cardiomyopathy (HOCM). The impact of the choice of optimization methods and other numerical settings, including fiber direction parameters, mesh size, initial parameters for optimization, and perturbations to optimal material parameters, is assessed using a rigorous sensitivity analysis. The performance of the current iFEA is compared against an assumed power-law-based pressure-volume relation, typically used for single-phase image acquisition.