Results 221 to 230 of about 395,369 (291)

MedVH: Toward Systematic Evaluation of Hallucination for Large Vision Language Models in the Medical Context

Advanced Intelligent Systems, EarlyView.
MedVH introduces the first comprehensive benchmark for diagnosing hallucinations in medical vision‐language models. Across six multitask evaluations, eight state‐of‐the‐art LVLMs reveal that domain‐tuned models, while strong on routine questions, hallucinate more than general models, raising serious concerns for real‐world clinical use.
Zishan Gu, Jiayuan Chen, Fenglin Liu, Changchang Yin, Ping Zhang   +4 more
wiley   +1 more source

Is Computing with Light All You Need? A Perspective on Codesign for Optical Artificial Intelligence and Scientific Computing

Advanced Intelligent Systems, EarlyView.
This perspective article considers what computations optical computing can and should enable. Focusing upon free‐space optical computing, it argues that a codesign approach whereby materials, devices, architectures, and algorithms are simultaneously optimized is needed.
Prasad P. Iyer, Gaurang R. Bhatt, Saaketh Desai, Elliot J. Fuller, Corinne M. Teeter, François Léonard, Craig M. Vineyard   +6 more
wiley   +1 more source

Learning Optimal Crowd Evacuation from Scratch Through Self‐Play

Advanced Intelligent Systems, EarlyView.
How would a superintelligent crowd behave in emergencies? This study develops an approach using multiagent deep reinforcement learning combined with self‐play to discover optimal evacuation strategies for pressure‐aware agents. The model learns behaviors such as queuing and zipper‐merging that significantly surpass traditional approaches in fatality ...
Mahdi Nasiri, Malte Cordts, Heinz Koeppl, Benno Liebchen   +3 more
wiley   +1 more source

Speech Recognition with Cochlea‐Inspired In‐Sensor Computing

Advanced Intelligent Systems, EarlyView.
Traditional speech recognition methods rely on software‐based feature extraction that introduces latency and high energy costs, making them unsuitable for low‐power devices. A proof‐of‐concept demonstration is provided of a bioinspired tonotopic sensor for speech recognition that mimics the human cochlea, using a spiral‐shaped elastic metamaterial. The
Paolo H. Beoletto, Gianluca Milano, Carlo Ricciardi, Federico Bosia, Antonio S. Gliozzi   +4 more
wiley   +1 more source

Revealing Patient Dissatisfaction With Health Care Resource Allocation in Multiple Dimensions Using Large Language Models and the International Classification of Diseases 11th Revision: Aspect-Based Sentiment Analysis.

J Med Internet Res
Li J, Yang Y, Mao C, Pang PC, Zhu Q, Xu D, Wang Y.   +6 more
europepmc   +1 more source

Enhancing Aspect-based Sentiment Analysis in Tourism Using Large Language Models and Positional Information [PDF]

Xu Chun, Mengmeng Wang, Yan Ren, Shaolin Zhu   +3 more
openalex   +1 more source

In situ ambient vibration modal analysis of saguaro cacti (Carnegiea gigantea)

American Journal of Botany, EarlyView.
Abstract Premise The structural and dynamic properties of columnar cacti are key inputs for stability analyses; however, no previous studies have been able to resolve these properties from full‐scale tests in situ. Methods I present an approach using non‐destructive ambient vibration data to measure the resonance properties (modal frequencies and mode ...
Jeffrey R. Moore
wiley   +1 more source

Aspect Based Sentiment Analysis using POS Tagging and TFIDF

, 2019
Kotagiri Srividya, A. Mary Sowjanya
openalex   +1 more source