Tag: #video

Google rolls out 'agentic video understanding' in Gemini Flash to cut token use and costs

Google adds 'agentic video understanding' to Gemini Flash models to analyze only relevant segments, cutting token use, cost and speeding long-video analysis.

$GOOGL, #google, #gemini, #ai, #video, #cloud

LAION publishes 10-million-hour open video dataset LAION-BVD; processed subsets available

LAION published LAION-BVD, a claimed 10-million-hour open video dataset for multimodal AI; processed subsets and pipelines are public, raw videos restricted.

#laion, #dataset, #ai, #video, #multimodal