Seeing People in 3D: A Sensor Showdown
![The research visualizes common data corruptions inherent in multi-sensor perception systems, specifically demonstrating how noise affects both camera imagery and [latex] 360^\circ [/latex] LiDAR point clouds- modalities labeled as C and L respectively-as captured within the JRDB dataset.](https://arxiv.org/html/2602.05538v1/x7.png)
A new study rigorously compares how different sensor technologies – cameras, LiDAR, and their combinations – perform in detecting people across real-world indoor and outdoor environments.
Science
Reading the Road: How Pedestrians React to Self-Driving Cars
![A spatial choice grid defines the available movement options for each pedestrian at a given time step, [latex] P_{nt} [/latex], enabling the system to model pedestrian navigation through discrete spatial possibilities.](https://arxiv.org/html/2602.05142v1/grid_def.jpg)
New research uses detailed movement analysis to understand how people adjust their walking speed when encountering autonomous vehicles.
Driven by Curiosity: AI Learns Best by Seeking What It Doesn’t Know
![The validation of Theorem 5.1 leveraged a discrete sandbox environment, with error bars denoting a margin of [latex] \pm 0.2 \pm 0.2 [/latex] standard deviations calculated across five independent trials.](https://arxiv.org/html/2602.06029v1/fig/consistency.png)
New research provides a theoretical framework demonstrating that an AI agent’s inherent ‘curiosity’ can guarantee optimal learning and decision-making in complex environments.
Seeing is Doing: Better Action Prediction with Visual Tracking
A new framework, VISTA, improves how robots understand visual cues and translate them into precise actions.
Simulating Life at Scale: A New Approach to Multiscale Cell Modeling
Researchers have developed a high-performance computing library that dramatically speeds up the simulation of complex biological systems, opening doors to virtual organ microenvironment studies.
Beyond Prediction: Reasoning for Trustworthy Recommendations
A new approach combines the power of collaborative filtering with large language models to not only predict what users want, but also explain why.
The Self-Evolving Algorithm: An AI That Rewrites Itself
Researchers have developed a novel framework where artificial intelligence agents collaboratively refine their own code, driving continuous improvement without constant human intervention.
Adapting to Change: A New Approach to Continuous Reasoning
Researchers have developed a novel system that dynamically adjusts its computational load based on how frequently data changes, enabling more efficient and responsive artificial intelligence.
When AI Turns Against Itself: The Rise of Proxy Attacks
New research reveals a concerning vulnerability in AI systems where agents can be exploited to circumvent safety protocols and deliver malicious instructions.
Stuck in a Rut: Why AI Agents Struggle to Explore
![Across all evaluated tasks, explore-exploit baselines consistently surpassed the performance of language models when operating under a query budget of [latex]N=48[/latex], demonstrating robustness to variations in parameter settings.](https://arxiv.org/html/2601.22345v1/x72.png)
A new evaluation benchmark reveals that current language models often fail to adequately explore interactive environments, leading to suboptimal decisions and a lack of adaptability.