Results 221 to 230 of about 1,593,960 (305)

LLM‐Based Multimodal Robotic Endoscope Control Framework for Enhancing Human–Robot Interaction in Minimally Invasive Surgery

Advanced Intelligent Systems, EarlyView.
An large language model‐powered multimodal framework is developed for robotic endoscope control. It integrates speech recognition and real‐time instrument tracking, achieving 89.47% command accuracy with ~1s latency for natural human–robot interaction in minimally invasive surgery.
Yisen Huang, Wing Yin Ng, Qi Dou, Xixin Wu, Shannon Melissa Chan, Yunhui Liu, Philip Wai Yan Chiu, Zheng Li   +7 more
wiley   +1 more source

IRA-YOLO: Inter-Class Relation Attention for PPE Detection in Industrial Scenes. [PDF]

Sensors (Basel)
Zhao R, Duan H, Liu X, Xu N, Du J, Zhou P.   +5 more
europepmc   +1 more source

Revealing Hidden Raman Signatures Through Attention‐Based Spectral Unmixing

Advanced Intelligent Systems, EarlyView.
Weak Raman signatures are recovered from background‐dominated spectra using a transformer‐based, reference‐free spectral unmixing AI framework. Self‐attention reconstructs substrate contribution directly from mixed data, enabling reliable extraction of previously inaccessible vibrational features.
Dmitriy A. Poteryayev, Pietro Novelli, Annalisa Coriolano, Riccardo Dettori, Valentina Tozzini, Fabio Beltram, Massimiliano Pontil, Antonio Rossi, Stiven Forti, Camilla Coletti   +9 more
wiley   +1 more source

Mapping the movie-watching brain with AI-derived semantics. [PDF]

Imaging Neurosci (Camb)
Li M.
europepmc   +1 more source

Resilient Multimodal Fusion for Robust Vulnerability Exploitability Prediction Under Data Degradation

Advanced Intelligent Systems, EarlyView.
Secure Fusion‐X harmonizes unstructured NVD descriptions with structured CVSS/CWE/CPE metadata via decision‐level fusion, overcoming the fragility of traditional unimodal models. Automated assessment of software vulnerability exploitability is essential for intelligent cyber defense, yet its effectiveness is often hindered by unstable, delayed, or ...
Mona Dolati, Seyedeh Leili Mirtaheri, Reza Shahbazian, Andrea Pugliese   +3 more
wiley   +1 more source

Task-Oriented Deep Joint Source-Channel Coding with Semantic-Aware Adaptive Quantization for Autonomous Driving. [PDF]

Sensors (Basel)
Wang X, Chen H, Sun Y, Li X, Xie M.
europepmc   +1 more source

Dual‐View Representation Learning for Deployable Humanoid Whole‐Body Control

Advanced Intelligent Systems, EarlyView.
Humanoid robots must learn control from limited onboard sensing, although simulation provides richer training‐only states. DVP learning pairs these views of each simulated motion and aligns their latent control features inside the policy‐learning loop. This direct, deployment‐aware supervision accelerates learning while the deployed policy relies only ...
Haolin Song, Wengang Zhou, Mingxiao Feng, Houqiang Li   +3 more
wiley   +1 more source

From Instruction to Inheritance: Scaling Robot Learning Through Knowledge Circulation

Advanced Intelligent Systems, EarlyView.
Robot learning moves from one‐way instruction toward knowledge circulation. Humans teach robots, robots teach one another, and robots share acquired skills with humans. This perspective connects teaching, embodiment, and learning architectures to explore how knowledge is inherited, adapted, and enriched across agents, opening a path toward scalable ...
Kento Kawaharazuka, Shunki Itadera, Kohei Honda, Takato Horii, Asako Kanezaki, Taisuke Kobayashi, Kenji Koide, Shuhei Kurita, Koshi Makihara, Kazuki Miyazawa, Masaki Murooka, Yuki Onishi, Satoshi Yagi, Satoshi Yamamori   +13 more
wiley   +1 more source

Model-based semantic distance reveals adaptive coordination of distinct cognitive systems in flexible knowledge retrieval. [PDF]

Nat Commun
Zhuang K, Liang X, Smallwood J, Jefferies E, Vatansever D.   +4 more
europepmc   +1 more source

DyCoDiff: Dynamic Collaborative Diffusion for Text‐Image‐Audio‐to‐Video

Advanced Intelligent Systems, EarlyView.
DyCoDiff enables text, image, and audio conditions to collaborate before temporal video generation. A multimodal condition fusion layer injects gated text‐centric fusion into the temporal branch, while collaborative losses align generated videos with semantic, visual, and dynamic audio cues for rhythm‐sensitive text‐image‐audio‐to‐video synthesis. Text‐
Youming Zhang, Wenmin Wang, Zhongheng Chen, Zhengyu Peng   +3 more
wiley   +1 more source