Vision-Language Models Stumble in Nutritional Wisdom
OmniFood-Bench sets a new standard for evaluating VLMs in dietary management. Despite advances, these models falter in important health-related tasks.
Read the full articleYou might also wanna read
Dutch modeling study finds plant-based protein swaps risk nutrient shortfalls in vulnerable age groups
A Dutch dietary modeling study found that replacing major animal-based protein sources with plant-based substitutes reduced protein, EAA, vi
news-medical.net·3d ago
LEVANTE-bench: Benchmark Reveals Partial Alignment Between Vision-Language Models and Children's Cognitive Abilities
Given the inherently multimodal nature of human experience, vision-language models (VLMs) hold substantial promise for modeling human cognit
arxiv.org·1mo ago
BabyVision Benchmark Reveals MLLMs Fail at Basic Visual Tasks That 3-Year-Olds Can Solve
While humans develop core visual skills long before acquiring language, contemporary Multimodal LLMs (MLLMs) still rely heavily on linguisti
arxiv.org·18d agoBiasing Food Value Construction
Conditions : Anorexia Nervosa Interventions : Behavioral: Training food attributes Sponsors : University of Chicago; National Institute of M
Community Dining, Walking, and Glucose in Older Adults
Conditions : Prediabetes Interventions : Behavioral: Subsidized Community Dining; Behavioral: Home Eating Sponsors : Yang Zhang Not yet recr
Suplementos vs. alimentos reales: cuál conviene y de qué depende, según la evidencia
Distingue los nutrientes aislados de la matriz de vitaminas, fibra y fitoquímicos presente en la dieta, y advierte sobre excesos, interaccio
Comments
Sign in to join the conversation.
No comments yet. Be the first.