Brief

Researchers have introduced a new benchmark suite designed to improve federated learning for medical image segmentation, specifically addressing the challenges posed by real-world label noise. This suite combines diverse noisy medical datasets with a comprehensive federated segmentation framework, offering realistic scenarios and noise-targeted evaluations. The goal is to facilitate systematic assessment and method selection for federated noisy label learning in medical imaging. AI

Researchers have introduced HAFMat, a novel framework designed to improve the estimation of physically based rendering (PBR) materials from single human images. This method addresses the inherent ambiguity in such estimations by employing a Multi-layer Adaptive Feature Fusion Mechanism. This mechanism adaptively integrates various guidance cues, including appearance, body geometry, and semantic information, at different stages of the decoding process. Experiments show that HAFMat achieves state-of-the-art results on both synthetic and real-world data for material estimation and subsequent relighting tasks. AI

Researchers have identified a new type of physical adversarial attack on vision systems, termed SLASH (Scratch-induced Lens Adversarial Streak Hijacking). This attack exploits small scratches on camera lenses or protective covers, which, when interacting with light sources, create structured streak artifacts that distort depth cues. The attack is persistent as the damage is fixed, but selective as it is triggered by specific scene conditions, leading to significant errors in monocular depth estimation and 3D object detection. AI