正在加载视频...

视频加载失败

加载此视频时出现问题。这可能是由于临时网络问题，或视频可能不可用。

Active Video Perception Iterative Evidence Seeking for Agentic Long Video Understanding

AK

526,440 subscribers

21,321 次观看 • 10 个月前 •via X (Twitter)

Anya Rossi• Live Now

Private livecam show

1 条评论

AK 的头像

AK10 个月前

discuss:

相关视频

Mode Seeking meets Mean Seeking for Fast Long Video Generation paper:

Mode Seeking meets Mean Seeking for Fast Long Video Generation paper:

AK

20,417 次观看 • 7 个月前

We’re introducing a new capability to our latest Gemini models: agentic video understanding. This allows developers to process long-form video content with more accuracy, while using up to 88% less tokens. See how it works 🧵

We’re introducing a new capability to our latest Gemini models: agentic video understanding. This allows developers to process long-form video content with more accuracy, while using up to 88% less tokens. See how it works 🧵

Google

793,471 次观看 • 1 个月前

Tired of high token costs for video workflows? Agentic video understanding gives Gemini the tools to dynamically search, scan, and inspect target video segments to achieve its goal. Building on agentic vision, Gemini now actively scans visual frames, audio, and transcripts instead of ingesting video at a static 1 FPS. This reduces token consumption by up to 88%, improves accuracy, and unlocks new ways to process long-form video. Tackle complex video workflows like: ▸ Sub-second retrieval: Catch split-second cuts missed at 1 FPS ▸ Long-form search: Query multi-hour videos without overflowing context ▸ Anomaly detection: Resample interesting windows at higher FPS ▸ Counting: Accurately track repeated actions and objects over time See how agentic video understanding transforms token efficiency in Gemini 3.7 Flash when compared with static video processing:

Tired of high token costs for video workflows? Agentic video understanding gives Gemini the tools to dynamically search, scan, and inspect target video segments to achieve its goal. Building on agentic vision, Gemini now actively scans visual frames, audio, and transcripts instead of ingesting video at a static 1 FPS. This reduces token consumption by up to 88%, improves accuracy, and unlocks new ways to process long-form video. Tackle complex video workflows like: ▸ Sub-second retrieval: Catch split-second cuts missed at 1 FPS ▸ Long-form search: Query multi-hour videos without overflowing context ▸ Anomaly detection: Resample interesting windows at higher FPS ▸ Counting: Accurately track repeated actions and objects over time See how agentic video understanding transforms token efficiency in Gemini 3.7 Flash when compared with static video processing:

Google for Developers

38,369 次观看 • 1 个月前

Agentic video understanding with Gemini 3.7 Flash is HUGE. Just reduced cost, less tokens, and higher accuracy. No biggie.

Agentic video understanding with Gemini 3.7 Flash is HUGE. Just reduced cost, less tokens, and higher accuracy. No biggie.

Shubham Saboo

79,642 次观看 • 1 个月前

Introducing Long Zhao, a Senior Research Scientist at Google, who worked to build VideoPrism: A Foundational Visual Encoder for Video Understanding. Read the blog to explore innovations in video understanding tasks and more →

Introducing Long Zhao, a Senior Research Scientist at Google, who worked to build VideoPrism: A Foundational Visual Encoder for Video Understanding. Read the blog to explore innovations in video understanding tasks and more →

Google AI

129,828 次观看 • 2 年前

This video is only 40s long . It will change your perception about President William Ruto.

This video is only 40s long . It will change your perception about President William Ruto.

Mr. Kinuthia Pius.

25,147 次观看 • 1 个月前

Excited to share our new blog: Scaling Video Training with Parallelism We discuss how parallelism helps scale long-video training systems, for both understanding and generation.

Excited to share our new blog: Scaling Video Training with Parallelism We discuss how parallelism helps scale long-video training systems, for both understanding and generation.

Yukang Chen

12,802 次观看 • 4 个月前

Attackers stopped working alone. They coordinate now—through agents. So, we built Project Perception: our new agentic security system. The video explains it better than a caption can.

Attackers stopped working alone. They coordinate now—through agents. So, we built Project Perception: our new agentic security system. The video explains it better than a caption can.

Microsoft Security

12,236 次观看 • 2 个月前

We’re launching agentic video understanding across our latest models: Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite! This new capability improves accuracy while cutting token consumption up to 88% and reducing costs by up to 66% for video analysis →

We’re launching agentic video understanding across our latest models: Gemini 3.7 Flash, 3.6 Flash, and 3.5 Flash-Lite! This new capability improves accuracy while cutting token consumption up to 88% and reducing costs by up to 66% for video analysis →

Google Cloud Tech

45,610 次观看 • 1 个月前

Video world models today have a very limited context length. Mode Seeking meets Mean Seeking (MMM) unlocks long-context, persistent video world models through a unified representation. 1/8 🧵

Video world models today have a very limited context length. Mode Seeking meets Mean Seeking (MMM) unlocks long-context, persistent video world models through a unified representation. 1/8 🧵

Gordon Wetzstein

43,945 次观看 • 7 个月前

StreamingVLM Real-Time Understanding for Infinite Video Streams

StreamingVLM Real-Time Understanding for Infinite Video Streams

AK

59,317 次观看 • 1 年前

Video breakdown: CVPR 2026 oral paper Shattering the efficiency bottleneck of long video understanding. Watch as we deconstruct the core innovations behind SpecTemp.

Video breakdown: CVPR 2026 oral paper Shattering the efficiency bottleneck of long video understanding. Watch as we deconstruct the core innovations behind SpecTemp.

FutureLivingLab

13,632 次观看 • 5 个月前

1. Breaking down + understanding a long video I uploaded the entire NBA dunk contest from last night and asked which dunk had the highest score. Gemini 1.5 was incredibly able to find the specific perfect 50 dunk and details from just its long context video understanding!

1. Breaking down + understanding a long video I uploaded the entire NBA dunk contest from last night and asked which dunk had the highest score. Gemini 1.5 was incredibly able to find the specific perfect 50 dunk and details from just its long context video understanding!

Rowan Cheung

101,170 次观看 • 2 年前

Check out this amazing combo using Gemma4 + Falcon Perception for video tracking! 1️⃣Give Gemma 4 video frames 2️⃣It describes what it sees 3️⃣Falcon Perception takes those descriptions, segments the objects, and tracks them across the video! The best part? All running locally!

Check out this amazing combo using Gemma4 + Falcon Perception for video tracking! 1️⃣Give Gemma 4 video frames 2️⃣It describes what it sees 3️⃣Falcon Perception takes those descriptions, segments the objects, and tracks them across the video! The best part? All running locally!

Google Gemma

84,378 次观看 • 6 个月前

Muse Spark 1.1 is here! Available today via API and on Of all the agentic use cases, I’m most excited about this one: going from a single video shot straight to task completion. 🚀 We’ve also polished and launched Computer Use and Video-Audio Perception for Muse Spark 1.1. Read more:

Muse Spark 1.1 is here! Available today via API and on Of all the agentic use cases, I’m most excited about this one: going from a single video shot straight to task completion. 🚀 We’ve also polished and launched Computer Use and Video-Audio Perception for Muse Spark 1.1. Read more:

Jiahui Yu

121,409 次观看 • 3 个月前

Random thoughts. Questions. How long would it take for y’all to get arrested if they had solid video evidence of your crimes?! 🤔🪓

Random thoughts. Questions. How long would it take for y’all to get arrested if they had solid video evidence of your crimes?! 🤔🪓

Cattleman🪓

37,948 次观看 • 9 个月前

This complete video will change your perception 👇

This complete video will change your perception 👇

DON “Tobechukwu” Ade 👑

32,451 次观看 • 3 个月前