LearningFlow: Automated Policy Learning Workflow for Urban Driving with Large Language Models

Recent advancements in reinforcement learning (RL) demonstrate the significant potential in autonomous driving. Despite this promise, challenges such as the manual design of reward functions and low sample efficiency in complex environments continue to impede the development of safe and effective driving policies. To tackle these issues, we introduce LearningFlow, an innovative automated policy learning workflow tailored to urban driving. This framework leverages the collaboration of multiple large language model (LLM) agents throughout the RL training process. LearningFlow includes a curriculum sequence generation process and a reward generation process, which work in tandem to guide the RL policy by generating tailored training curricula and reward functions. Particularly, each process is supported by an analysis agent that evaluates training progress and provides critical insights to the generation agent. Through the collaborative efforts of these LLM agents, LearningFlow automates policy learning across a series of complex driving tasks, and it significantly reduces the reliance on manual reward function design while enhancing sample efficiency. Comprehensive experiments are conducted in the high-fidelity CARLA simulator, along with comparisons with other existing methods, to demonstrate the efficacy of our proposed approach. The results demonstrate that LearningFlow excels in generating rewards and curricula. It also achieves superior performance and robust generalization across various driving tasks, as well as commendable adaptation to different RL algorithms.

Local Reactive Control for Mobile Manipulators with Whole-Body Safety in Complex Environments

Mobile manipulators typically encounter significant challenges in navigating narrow, cluttered environments due to their high-dimensional state spaces and complex kinematics. While reactive methods excel in dynamic settings, they struggle to efficiently incorporate complex, coupled constraints across the entire state space. In this work, we present a novel local reactive controller that reformulates the time-domain single-step problem into a multi-step optimization problem in the spatial domain, leveraging the propagation of a serial kinematic chain. This transformation facilitates the formulation of customized, decoupled link-specific constraints, which is further solved efficiently with augmented Lagrangian differential dynamic programming (AL-DDP). Our approach naturally absorbs spatial kinematic propagation in the forward pass and processes all link-specific constraints simultaneously during the backward pass, enhancing both constraint management and computational efficiency. Notably, in this framework, we formulate collision avoidance constraints for each link using accurate geometric models with extracted free regions, and this improves the maneuverability of the mobile manipulator in narrow, cluttered spaces. Experimental results showcase significant improvements in safety, efficiency, and task completion rates. These findings underscore the robustness of the proposed method, particularly in narrow, cluttered environments where conventional approaches could falter. The open-source project can be found at https://github.com/Chunx1nZHENG/MM-with-Whole-Body-Safety-Release.git.

UDMC: Unified Decision-Making and Control Framework for Urban Autonomous Driving with Motion Prediction of Traffic Participants

Current autonomous driving systems often struggle to balance decision-making and motion control while ensuring safety and traffic rule compliance, especially in complex urban environments. Existing methods may fall short due to separate handling of these functionalities, leading to inefficiencies and safety compromises. To address these challenges, we introduce UDMC, an interpretable and unified Level 4 autonomous driving framework. UDMC integrates decision-making and motion control into a single optimal control problem (OCP), considering the dynamic interactions with surrounding vehicles, pedestrians, road lanes, and traffic signals. By employing innovative potential functions to model traffic participants and regulations, and incorporating a specialized motion prediction module, our framework enhances on-road safety and rule adherence. The integrated design allows for real-time execution of flexible maneuvers suited to diverse driving scenarios. High-fidelity simulations conducted in CARLA exemplify the framework's computational efficiency, robustness, and safety, resulting in superior driving performance when compared against various baseline models. Our open-source project is available at https://github.com/henryhcliu/udmc_carla.git.

LSAQ: Layer-Specific Adaptive Quantization for Large Language Model Deployment

As large language models (LLMs) demonstrate exceptional performance across various domains, the deployment of these models on edge devices has emerged as a new trend. Quantization techniques, which reduce the size and memory footprint of LLMs, are effective for enabling deployment on resource-constrained edge devices. However, existing one-size-fits-all quantization methods often fail to dynamically adjust the memory consumption of LLMs based on specific hardware characteristics and usage scenarios. To address this limitation, we propose LSAQ (Layer-Specific Adaptive Quantization), a system for adaptive quantization and dynamic deployment of LLMs based on layer importance. LSAQ evaluates layer importance by constructing top-k token sets from the inputs and outputs of each layer and calculating their Jaccard coefficient. Using this evaluation, the system adaptively adjusts quantization strategies in real time according to the resource availability of edge devices, assigning different precision levels to layers of varying importance. This approach significantly reduces the storage requirements of LLMs while maintaining model performance, enabling efficient deployment across diverse hardware platforms and usage scenarios.

Efficient MedSAMs: Segment Anything in Medical Images on Laptop

Promptable segmentation foundation models have emerged as a transformative approach to addressing the diverse needs in medical images, but most existing models require expensive computing, posing a big barrier to their adoption in clinical practice. In this work, we organized the first international competition dedicated to promptable medical image segmentation, featuring a large-scale dataset spanning nine common imaging modalities from over 20 different institutions. The top teams developed lightweight segmentation foundation models and implemented an efficient inference pipeline that substantially reduced computational requirements while maintaining state-of-the-art segmentation accuracy. Moreover, the post-challenge phase advanced the algorithms through the design of performance booster and reproducibility tasks, resulting in improved algorithms and validated reproducibility of the winning solution. Furthermore, the best-performing algorithms have been incorporated into the open-source software with a user-friendly interface to facilitate clinical adoption. The data and code are publicly available to foster the further development of medical image segmentation foundation models and pave the way for impactful real-world applications.

CALMM-Drive: Confidence-Aware Autonomous Driving with Large Multimodal Model

Decision-making and motion planning are pivotal in ensuring the safety and efficiency of Autonomous Vehicles (AVs). Existing methodologies typically adopt two paradigms: decision then planning or generation then scoring. However, the former often struggles with misalignment between decisions and planning, while the latter encounters significant challenges in integrating short-term operational utility with long-term tactical efficacy. To address these issues, we introduce CALMM-Drive, a novel Confidence-Aware Large Multimodal Model (LMM) empowered Autonomous Driving framework. Our approach employs Top-K confidence elicitation, which facilitates the generation of multiple candidate decisions along with their confidence levels. Furthermore, we propose a novel planning module that integrates a diffusion model for trajectory generation and a hierarchical refinement process to find the optimal path. This framework enables the selection of the best plan accounting for both low-level solution quality and high-level tactical confidence, which mitigates the risks of one-shot decisions and overcomes the limitations induced by short-sighted scoring mechanisms. Comprehensive evaluations in nuPlan closed-loop simulation environments demonstrate the effectiveness of CALMM-Drive in achieving reliable and flexible driving performance, showcasing a significant advancement in the integration of uncertainty in LMM-empowered AVs. The code will be released upon acceptance.

Integrating Decision-Making Into Differentiable Optimization Guided Learning for End-to-End Planning of Autonomous Vehicles

We address the decision-making capability within an end-to-end planning framework that focuses on motion prediction, decision-making, and trajectory planning. Specifically, we formulate decision-making and trajectory planning as a differentiable nonlinear optimization problem, which ensures compatibility with learning-based modules to establish an end-to-end trainable architecture. This optimization introduces explicit objectives related to safety, traveling efficiency, and riding comfort, guiding the learning process in our proposed pipeline. Intrinsic constraints resulting from the decision-making task are integrated into the optimization formulation and preserved throughout the learning process. By integrating the differentiable optimizer with a neural network predictor, the proposed framework is end-to-end trainable, aligning various driving tasks with ultimate performance goals defined by the optimization objectives. The proposed framework is trained and validated using the Waymo Open Motion dataset. The open-loop testing reveals that while the planning outcomes using our method do not always resemble the expert trajectory, they consistently outperform baseline approaches with improved safety, traveling efficiency, and riding comfort. The closed-loop testing further demonstrates the effectiveness of optimizing decisions and improving driving performance. Ablation studies demonstrate that the initialization provided by the learning-based prediction module is essential for the convergence of the optimizer as well as the overall driving performance.

Synergizing Decision Making and Trajectory Planning Using Two-Stage Optimization for Autonomous Vehicles

This paper introduces a local planner that synergizes the decision making and trajectory planning modules towards autonomous driving. The decision making and trajectory planning tasks are jointly formulated as a nonlinear programming problem with an integrated objective function. However, integrating the discrete decision variables into the continuous trajectory optimization leads to a mixed-integer programming (MIP) problem with inherent nonlinearity and nonconvexity. To address the challenge in solving the problem, the original problem is decomposed into two sub-stages, and a two-stage optimization (TSO) based approach is presented to ensure the coherence in outcomes for the two stages. The optimization problem in the first stage determines the optimal decision sequence that acts as an informed initialization. With the outputs from the first stage, the second stage necessitates the use of a high-fidelity vehicle model and strict enforcement of the collision avoidance constraints as part of the trajectory planning problem. We evaluate the effectiveness of our proposed planner across diverse multi-lane scenarios. The results demonstrate that the proposed planner simultaneously generates a sequence of optimal decisions and the corresponding trajectory that significantly improves driving performance in terms of driving safety and traveling efficiency as compared to alternative methods. Additionally, we implement the closed-loop simulation in CARLA, and the results showcase the effectiveness of the proposed planner to adapt to changing driving situations with high computational efficiency.

MOSABench: Multi-Object Sentiment Analysis Benchmark for Evaluating Multimodal Large Language Models Understanding of Complex Image

Multimodal large language models (MLLMs) have shown remarkable progress in high-level semantic tasks such as visual question answering, image captioning, and emotion recognition. However, despite advancements, there remains a lack of standardized benchmarks for evaluating MLLMs performance in multi-object sentiment analysis, a key task in semantic understanding. To address this gap, we introduce MOSABench, a novel evaluation dataset designed specifically for multi-object sentiment analysis. MOSABench includes approximately 1,000 images with multiple objects, requiring MLLMs to independently assess the sentiment of each object, thereby reflecting real-world complexities. Key innovations in MOSABench include distance-based target annotation, post-processing for evaluation to standardize outputs, and an improved scoring mechanism. Our experiments reveal notable limitations in current MLLMs: while some models, like mPLUG-owl and Qwen-VL2, demonstrate effective attention to sentiment-relevant features, others exhibit scattered focus and performance declines, especially as the spatial distance between objects increases. This research underscores the need for MLLMs to enhance accuracy in complex, multi-object sentiment analysis tasks and establishes MOSABench as a foundational tool for advancing sentiment analysis capabilities in MLLMs.

Robotic Control Optimization Through Kernel Selection in Safe Bayesian Optimization

Control system optimization has long been a fundamental challenge in robotics. While recent advancements have led to the development of control algorithms that leverage learning-based approaches, such as SafeOpt, to optimize single feedback controllers, scaling these methods to high-dimensional complex systems with multiple controllers remains an open problem. In this paper, we propose a novel learning-based control optimization method, which enhances the additive Gaussian process-based Safe Bayesian Optimization algorithm to efficiently tackle high-dimensional problems through kernel selection. We use PID controller optimization in drones as a representative example and test the method on Safe Control Gym, a benchmark designed for evaluating safe control techniques. We show that the proposed method provides a more efficient and optimal solution for high-dimensional control optimization problems, demonstrating significant improvements over existing techniques.