PulseAugur / Brief
EN
LIVE 11:06:47

Brief

last 24h
[2/2] 224 sources

Multi-source AI news clustered, deduplicated, and scored 0–100 across authority, cluster strength, headline signal, and time decay.

  1. TOOL · arXiv cs.AI English(EN) ·

    Ghost Attractor Networks: Basin-Structured Dynamical Decoders for Closed-Loop Sequential Generation

    Researchers have introduced Ghost Attractor Networks (GANs), a novel dynamical decoder designed to improve sequential generation efficiency and control in large-scale models. GANs utilize a learned potential with a basin-attractor structure to enable closed-loop control, such as phase-conditioned action generation and cross-step latent carry-over. Empirically, a GAN model demonstrated a significant reduction in parameters and latency compared to a Diffusion Transformer while achieving comparable or superior accuracy on robotic action decoding tasks and closed-loop benchmarks. AI

    IMPACT Introduces a more efficient and controllable method for sequential generation, potentially impacting robotics and other generative AI applications.

  2. RESEARCH · arXiv cs.LG English(EN) · · [2 sources]

    Where Will They Go? Modelling Multimodal Pedestrian Manoeuvres from Ego-centric Videos

    Researchers have developed a new framework called MMPM to improve pedestrian trajectory prediction from ego-centric videos. This model addresses the challenge of multimodal pedestrian behavior by separately modeling distinct modes, such as crossing or not crossing the road. The MMPM framework includes a behavior-aware Pedestrian Interaction Module (PIM) and a CVAE-based Mode-aware Trajectory Predictor (MTP), which collectively capture complex interactions and intentions. Experiments on PIE and JAAD datasets demonstrate that MMPM outperforms existing state-of-the-art methods and can be integrated with other frameworks like BiTrap-NP and SGNet-ED. AI

    IMPACT Enhances the accuracy of predicting pedestrian movements in complex urban environments, potentially improving autonomous navigation and safety systems.