Daniel Yaminsمشاهده پروفایل
دانشیار
Associate Professor Daniel Yamins holds dual appointments in Stanford University's Department of Psychology and Department of Computer Science. His research focuses on the intersection of neuroscience and artificial intelligence, particularly in understanding how biological and artificial systems process visual information and learn physical dynamics. He leads the NeuroAILab, which develops computational models to explain neural processes in perception, cognition, and development. Yamins earned his Ph.D. in Applied Mathematics from Harvard University in 2008. His work integrates machine learning, cognitive science, and neuroimaging to bridge gaps between biological and artificial intelligence. Key projects include the BabyView dataset, which captures infants' visual experiences, and Physion++, a benchmark for evaluating physical scene understanding in humans and machines. His research interests span object perception, counterfactual reasoning, 3D scene understanding, and developmental robotics. Recent work emphasizes self-supervised learning, world modeling, and the design of image-computable neural network architectures that mirror cortical processing. Yamins has contributed to open-source platforms like ThreedWorld for physically realistic simulation and Hyperopt for hyperparameter optimization. His grants and collaborations include NSF-funded research on neural mechanisms of physical scene understanding and NIH-backed studies on neurodevelopmental processes. He actively bridges theoretical neuroscience with practical AI applications, advocating for benchmark-driven model evaluation aligned with biological plausibility.
