Mapping Code’s Structure for Smarter AI Agents
Formally describing a codebase’s architecture can significantly improve the efficiency and consistency of AI coding assistants.
Science
Seeing Beyond Obstructions: Vision-Based Robot Control with Image Inpainting
This review details a markerless vision system that enables robotic manipulators to reliably detect keypoints and perform control tasks even in cluttered or partially obscured environments.
AI Tackles Fiber Optics: A New Approach to Formula Derivation
Researchers have successfully employed artificial intelligence to reconstruct and extend key formulas governing light propagation in fiber optic cables, opening doors for faster and more accurate modeling.
Seeing the World in Full Circle: A New Vision for Robotics
![RobotPan establishes a real-time embodied perception system by fusing data from six cameras and LiDAR, enabling the prediction of metric-scaled [latex]3D[/latex] Gaussians from sparse multi-view observations and facilitating applications such as surround-view rendering, novel view synthesis, depth estimation, and sparse-view dense reconstruction through a jointly optimized framework prioritizing geometric consistency, compact representation, and real-time performance.](https://arxiv.org/html/2604.13476v1/figures/cover.jpg)
Researchers have developed a novel surround-view system that allows robots to perceive and reconstruct 3D environments in real-time using sparse visual data.
Beyond Implementation: AI Agents Now Refine Published Algorithms
New research demonstrates that AI-powered tools can systematically improve existing algorithm implementations, shifting the focus for researchers towards validation and direction.
Learning to Move: Diffusion Models Unlock Robust Robot Control
New research demonstrates that diffusion models can significantly improve a robot’s ability to learn and adapt to new movements and environments.
DIY Autonomy: Building a Self-Driving Lab with Off-the-Shelf Parts
![An affordable, closed-loop system iteratively refines voltage commands to an LED array-guided by traversal search, Bayesian optimization utilizing a probabilistic surrogate model [latex]\mu(x)[/latex], or a deep learning neural network-and compares the resulting discrete spectrum detected by a multichannel light sensor to a user-defined target, effectively creating a self-driving optical experiment.](https://arxiv.org/html/2604.13139v1/Figures/Fig1.png)
Researchers demonstrate a low-cost, internet-connected platform for automated physics experiments, paving the way for hands-on machine learning education and autonomous discovery.
Bringing Video to Life: Reconstructing Human Interaction in 3D
![SocialMirror establishes a framework that refines outputs through the interplay of semantic understanding-derived from vision-language annotations-and geometric constraints, leveraging [latex]Trans Block[/latex] components to achieve nuanced control.](https://arxiv.org/html/2604.13581v1/x1.png)
Researchers have developed a new AI framework that accurately recreates realistic 3D human poses and behaviors from standard video footage.
Grading on a Curve: AI Steps into Peer Review
![A study of peer review responses indicates that automated assessments frequently surpassed human evaluations across multiple quality criteria-a preference notably stronger among authors-and, despite exceeding initial expectations, these AI reviews demonstrated both unique strengths in identifying nuanced issues and predictable limitations in overlooking others, suggesting a complementary rather than substitutive role alongside human judgment in the evolving landscape of scholarly evaluation-all findings supported by statistically significant results [latex]\alpha = 0.01[/latex].](https://arxiv.org/html/2604.13940v1/x4.png)
A large-scale pilot program at a leading AI conference demonstrates the growing potential of artificial intelligence to assist with the critical process of scientific evaluation.
Easing the Path: Predicting Pedestrian Comfort Around Robots
![The study defines key spatial kinematic variables-minimum relative Euclidean distance [latex]D_{min}[/latex], lateral distance at passing [latex]D_{lat}[/latex], maximum trajectory curvature ρ, and distance at the point of temporal proximity [latex]D_{T_{P}}[/latex]-to characterize the robot’s navigation around pedestrians, acknowledging that each parameter contributes to a predictable geometry of potential failure in dynamic human-robot interaction.](https://arxiv.org/html/2604.13677v1/x4.png)
New research quantifies how robot movement impacts pedestrian comfort levels, paving the way for more natural and safer human-robot interactions.