
  <rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
    <channel>
      <title>Chaos and Order</title>
      <link>https://www.youngju.dev/blog</link>
      <description>천천히 올바르게. AI Researcher &amp; DevOps Engineer Youngju&#39;s tech blog. GPU/CUDA, LLM, MLOps, Kubernetes AI workloads, distributed training, and data engineering.</description>
      <language>ko</language>
      <managingEditor>fjvbn2003@gmail.com (Youngju Kim)</managingEditor>
      <webMaster>fjvbn2003@gmail.com (Youngju Kim)</webMaster>
      <lastBuildDate>Sat, 16 May 2026 00:00:00 GMT</lastBuildDate>
      <atom:link href="https://www.youngju.dev/tags/twelve-labs/feed.xml" rel="self" type="application/rss+xml"/>
      
  <item>
    <guid>https://www.youngju.dev/blog/culture/2026-05-16-ai-video-search-indexing-2026-twelve-labs-pinecone-multimodal-roboflow-video-cloudflare-stream-mux-asset-metadata-google-video-intelligence-deep-dive.en</guid>
    <title>AI Video Search &amp; Indexing 2026 Complete Guide — Twelve Labs, Pinecone Multimodal, Roboflow Video Inference, Cloudflare Stream, Mux Asset Metadata, Google Video Intelligence, AWS Rekognition Video, Azure Video Indexer In-Depth Analysis</title>
    <link>https://www.youngju.dev/blog/culture/2026-05-16-ai-video-search-indexing-2026-twelve-labs-pinecone-multimodal-roboflow-video-cloudflare-stream-mux-asset-metadata-google-video-intelligence-deep-dive.en</link>
    <description>In 2026, video stopped being a thing you watch and became data you search, summarize, and cite. Thousands of meeting recording hours, petabytes of CCTV footage, scene-level search across content libraries, and multimodal RAG all share the same infrastructure. This guide walks through the entire 2026 landscape — Twelve Labs Pegasus 1.2 and Marengo 2.7, multimodal embeddings like Google SigLIP2, Cohere Embed v3, Voyage Multimodal and Nomic Embed Multimodal; vector databases including Pinecone Multimodal, Weaviate multi2vec-clip, Qdrant and Milvus; hyperscaler APIs like Google Video Intelligence, AWS Rekognition Video and Azure Video Indexer; object detection with Roboflow Video, YOLO and NVIDIA DeepStream; foundation models like Sora, Veo 2, Runway Gen-3, Gemini 2.0, GPT-4o, Claude 3.5 and InternVL; asset metadata with Mux, Cloudflare Stream and JW Player; real use cases across meetings, security, content, e-commerce and live; plus Korean and Japanese vendors and storage cost economics.</description>
    <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
    <author>fjvbn2003@gmail.com (Youngju Kim)</author>
    <category>ai-video-search</category><category>video-indexing</category><category>twelve-labs</category><category>pinecone-multimodal</category><category>roboflow-video</category><category>cloudflare-stream</category><category>mux</category><category>google-video-intelligence</category><category>aws-rekognition</category><category>azure-video-indexer</category><category>english</category>
  </item>

  <item>
    <guid>https://www.youngju.dev/blog/culture/2026-05-16-ai-video-search-indexing-2026-twelve-labs-pinecone-multimodal-roboflow-video-cloudflare-stream-mux-asset-metadata-google-video-intelligence-deep-dive.ja</guid>
    <title>AI 動画検索 &amp; インデキシング 2026 完全ガイド - Twelve Labs · Pinecone Multimodal · Roboflow Video Inference · Cloudflare Stream · Mux Asset Metadata · Google Video Intelligence · AWS Rekognition Video · Azure Video Indexer 徹底解析</title>
    <link>https://www.youngju.dev/blog/culture/2026-05-16-ai-video-search-indexing-2026-twelve-labs-pinecone-multimodal-roboflow-video-cloudflare-stream-mux-asset-metadata-google-video-intelligence-deep-dive.ja</link>
    <description>2026年の動画はもはや「観るもの」ではなく「検索・要約・引用」できるデータになった。会議録画の数千時間、CCTVのペタバイト、コンテンツライブラリのシーン検索、マルチモーダルRAGがすべて同じインフラを共有する。本稿は、Twelve LabsのPegasus 1.2 / Marengo 2.7、Google SigLIP2やCohere Embed v3・Voyage Multimodal・Nomic Embed Multimodalなどのマルチモーダル埋め込み、Pinecone Multimodal・Weaviate multi2vec-clip・Qdrant・Milvusなどのベクトル DB、Google Video Intelligence・AWS Rekognition・Azure Video Indexerなどのハイパースケーラ API、Roboflow Video Inference・YOLO・NVIDIA DeepStreamなどの物体検出、Sora・Veo 2・Runway Gen-3・Gemini 2.0・GPT-4o・Claude 3.5・InternVLなどのファウンデーションモデル、Mux Asset Metadata・Cloudflare Stream・JW Playerなどのアセット メタデータ、会議・CCTV・コンテンツ・EC・ライブの実用ケース、韓国・日本のローカルベンダー、ストレージ・帯域コストまでをひとつの流れで整理する。</description>
    <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
    <author>fjvbn2003@gmail.com (Youngju Kim)</author>
    <category>ai-video-search</category><category>video-indexing</category><category>twelve-labs</category><category>pinecone-multimodal</category><category>roboflow-video</category><category>cloudflare-stream</category><category>mux</category><category>google-video-intelligence</category><category>aws-rekognition</category><category>azure-video-indexer</category><category>日本語</category>
  </item>

  <item>
    <guid>https://www.youngju.dev/blog/culture/2026-05-16-ai-video-search-indexing-2026-twelve-labs-pinecone-multimodal-roboflow-video-cloudflare-stream-mux-asset-metadata-google-video-intelligence-deep-dive</guid>
    <title>AI 비디오 검색 &amp; 인덱싱 2026 완벽 가이드 - Twelve Labs · Pinecone Multimodal · Roboflow Video Inference · Cloudflare Stream · Mux Asset Metadata · Google Video Intelligence · AWS Rekognition Video · Azure Video Indexer 심층 분석</title>
    <link>https://www.youngju.dev/blog/culture/2026-05-16-ai-video-search-indexing-2026-twelve-labs-pinecone-multimodal-roboflow-video-cloudflare-stream-mux-asset-metadata-google-video-intelligence-deep-dive</link>
    <description>2026년의 비디오는 더 이상 보기 위한 것이 아니라 검색·요약·인용을 위한 데이터다. 회의 녹화 수천 시간, CCTV 페타바이트, 콘텐츠 라이브러리의 장면 검색, 멀티모달 RAG가 같은 인프라를 공유한다. Twelve Labs의 Pegasus 1.2/Marengo 2.7, Google SigLIP2, Cohere Embed v3·Voyage Multimodal·Nomic Embed Multimodal 같은 임베딩, Pinecone Multimodal·Weaviate multi2vec-clip·Qdrant·Milvus 같은 벡터 DB, Google Video Intelligence·AWS Rekognition·Azure Video Indexer 같은 하이퍼스케일러 API, Roboflow Video Inference·YOLO·NVIDIA DeepStream 같은 객체 검출, Sora·Veo 2·Runway Gen-3·Gemini 2.0·GPT-4o·Claude 3.5·InternVL 같은 파운데이션 모델, Mux Asset Metadata·Cloudflare Stream·JW Player 같은 에셋 메타데이터, 회의·CCTV·콘텐츠·이커머스·라이브 같은 활용처, 한국·일본의 자국 사업자까지 한 흐름으로 묶어 정리한다.</description>
    <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
    <author>fjvbn2003@gmail.com (Youngju Kim)</author>
    <category>ai-video-search</category><category>video-indexing</category><category>twelve-labs</category><category>pinecone-multimodal</category><category>roboflow-video</category><category>cloudflare-stream</category><category>mux</category><category>google-video-intelligence</category><category>aws-rekognition</category><category>azure-video-indexer</category>
  </item>

    </channel>
  </rss>
