New dataset STSBench advances modeling of primate visual cortex dorsal stream

Researchers have introduced STSBench, a new large-scale dataset designed to advance the modeling of neuronal activity within the dorsal stream of the primate visual cortex. This dataset comprises recordings from over 2,000 neurons in the superior temporal sulcus, significantly expanding upon previous dorsal stream datasets. STSBench aims to facilitate the benchmarking of encoding models for dorsal stream responses and enable the reconstruction of visual input from neural activity, addressing a gap in current research where convolutional neural networks have been more effectively applied to the ventral stream. AI

RANK_REASON The item describes a new dataset for research in neuroscience, specifically for modeling neuronal activity. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.CV →

• arXiv
• CNNS
• convolutional neural network
• Primate Visual Cortex
• rhesus macaque
• STSBench
• superior temporal sulcus

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →