Google Unveils Agentic Vision, Empowering Gemini 3 Flash with Active Visual Reasoning Capabilities - AI

7 x 24 Track global technological trends

Hot Topic

Day

News Topic

Google Unveils Agentic Vision, Empowering Gemini 3 Flash with Active Visual Reasoning Capabilities

1 week ago / Read about 0 minute

Author：小编

Recently, Google rolled out the 'Agentic Vision' capability within its Gemini 3 Flash model, marking a significant leap from conventional static image recognition to dynamic, investigative visual comprehension. This innovative feature functions through an iterative 'think-act-observe' cycle, seamlessly integrating visual reasoning with code execution. Such integration allows the model to autonomously navigate and analyze images, resulting in a notable 5% to 10% boost in accuracy when tackling intricate visual tasks.

Previous page：OpenAI's Pre-Funding Valuation Could Soar to $730 ...

Next page：Chief Data Officer: Occupying a 'Novel C-Suite Rol...

Return to List

Hot Reading

1 day ago

DJI Osmo Pocket 4 Camera Leaks: 6K, AI Tracking, and Pro Features Revealed

2 day ago

Intel clawed back CPU market share from AMD in the Steam Hardware survey for the first time in months

2 day ago

Tesla hit with another wrongful death suit over its electronic door handles

2 day ago

Steam Machine and Steam Frame delays are the latest product of the RAM crisis