
  <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>Fri, 15 May 2026 00:00:00 GMT</lastBuildDate>
      <atom:link href="https://www.youngju.dev/tags/luma-ai/feed.xml" rel="self" type="application/rss+xml"/>
      
  <item>
    <guid>https://www.youngju.dev/blog/culture/2026-05-15-ai-video-generation-2026-sora-2-veo-3-kling-2-hailuo-runway-gen-4-luma-hunyuan-video-deep-dive.en</guid>
    <title>AI Video Generation 2026 — Sora 2 / Veo 3 / Kling 2 / Hailuo / Runway Gen-4 / Luma Ray 2 / HunyuanVideo Deep Dive</title>
    <link>https://www.youngju.dev/blog/culture/2026-05-15-ai-video-generation-2026-sora-2-veo-3-kling-2-hailuo-runway-gen-4-luma-hunyuan-video-deep-dive.en</link>
    <description>Two years after Sora 1 stunned the film industry in Feb 2024, the AI video generation landscape of May 2026 has hardened around three camps — closed-source SOTA (Sora 2, Veo 3, Kling 2, Hailuo), industry-standard video tools (Runway Gen-4, Luma Ray 2, Pika 2), and a credible open-weight tier (HunyuanVideo, LTX-Video, Wan 2.1, Open-Sora). Add Lyria/Suno/Udio for music and ElevenLabs SFX and you finally have a unified workflow. This deep dive maps it in 14 chapters.</description>
    <pubDate>Fri, 15 May 2026 00:00:00 GMT</pubDate>
    <author>fjvbn2003@gmail.com (Youngju Kim)</author>
    <category>ai-video</category><category>video-generation</category><category>sora</category><category>veo</category><category>kling</category><category>hailuo</category><category>pika</category><category>runway</category><category>luma-ai</category><category>hunyuanvideo</category><category>ltx-video</category><category>wan</category><category>open-sora</category><category>dit</category><category>2026</category><category>deep-dive</category><category>english</category>
  </item>

  <item>
    <guid>https://www.youngju.dev/blog/culture/2026-05-15-ai-video-generation-2026-sora-2-veo-3-kling-2-hailuo-runway-gen-4-luma-hunyuan-video-deep-dive.ja</guid>
    <title>AI動画生成 2026 — Sora 2 / Veo 3 / Kling 2 / Hailuo / Runway Gen-4 / Luma Ray 2 / HunyuanVideo 徹底ガイド</title>
    <link>https://www.youngju.dev/blog/culture/2026-05-15-ai-video-generation-2026-sora-2-veo-3-kling-2-hailuo-runway-gen-4-luma-hunyuan-video-deep-dive.ja</link>
    <description>2024年2月のSora 1デモが映像業界をひっくり返してから2年。Sora 2が正式リリースされ、Veo 3が音声と会話を同期生成し、Kling 2が中国発の最強動画モデルとなり、HunyuanVideoがオープンウェイトの初の競合になった。2026年5月時点のAI動画生成の地形 — クローズド/オープン/音声・音楽連携/日韓研究まで14章で整理。</description>
    <pubDate>Fri, 15 May 2026 00:00:00 GMT</pubDate>
    <author>fjvbn2003@gmail.com (Youngju Kim)</author>
    <category>ai-video</category><category>video-generation</category><category>sora</category><category>veo</category><category>kling</category><category>hailuo</category><category>pika</category><category>runway</category><category>luma-ai</category><category>hunyuanvideo</category><category>ltx-video</category><category>wan</category><category>open-sora</category><category>dit</category><category>2026</category><category>deep-dive</category><category>日本語</category>
  </item>

  <item>
    <guid>https://www.youngju.dev/blog/culture/2026-05-15-ai-video-generation-2026-sora-2-veo-3-kling-2-hailuo-runway-gen-4-luma-hunyuan-video-deep-dive</guid>
    <title>AI 비디오 생성 2026 — Sora 2 / Veo 3 / Kling 2 / Hailuo / Runway Gen-4 / Luma Ray 2 / HunyuanVideo 심층 가이드</title>
    <link>https://www.youngju.dev/blog/culture/2026-05-15-ai-video-generation-2026-sora-2-veo-3-kling-2-hailuo-runway-gen-4-luma-hunyuan-video-deep-dive</link>
    <description>2024년 2월 Sora 1의 60초 광고가 영상 업계를 뒤집은 지 2년. Sora 2가 정식 출시되고, Google Veo 3는 합성 오디오와 대화까지 만들고, Kling 2는 중국발 최강 영상 모델로 자리잡고, HunyuanVideo가 오픈웨이트 경쟁자를 처음으로 만들었다. 2026년 5월 시점의 AI 비디오 생성 지형도 — 폐쇄형 / 오픈웨이트 / 오디오·음악 결합 / 한국·일본 연구까지 14개 장으로 정리.</description>
    <pubDate>Fri, 15 May 2026 00:00:00 GMT</pubDate>
    <author>fjvbn2003@gmail.com (Youngju Kim)</author>
    <category>ai-video</category><category>video-generation</category><category>sora</category><category>veo</category><category>kling</category><category>hailuo</category><category>pika</category><category>runway</category><category>luma-ai</category><category>hunyuanvideo</category><category>ltx-video</category><category>wan</category><category>open-sora</category><category>dit</category><category>2026</category><category>deep-dive</category>
  </item>

  <item>
    <guid>https://www.youngju.dev/blog/culture/2026-05-16-3d-gaussian-splatting-photogrammetry-2026-nerfstudio-postshot-luma-genie-polycam-apple-object-capture-realitycapture-deep-dive.en</guid>
    <title>3D Gaussian Splatting &amp; Photogrammetry 2026 Deep Dive - Nerfstudio · Postshot · Luma Genie · Polycam · Apple Object Capture · RealityCapture Field Guide</title>
    <link>https://www.youngju.dev/blog/culture/2026-05-16-3d-gaussian-splatting-photogrammetry-2026-nerfstudio-postshot-luma-genie-polycam-apple-object-capture-realitycapture-deep-dive.en</link>
    <description>In 2023, an INRIA paper closed the NeRF era. By 2026, a hundred photos and a desktop are enough to bake photorealistic 3D. This guide walks the entire 3DGS + photogrammetry stack: Nerfstudio 1.4 with Splatfacto, Postshot 2.0, Luma Genie 2, Polycam, Apple Object Capture on macOS Sequoia, the now-free RealityCapture 1.5, Scaniverse 4, and Meshroom — from capture to web viewer in one breath.</description>
    <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
    <author>fjvbn2003@gmail.com (Youngju Kim)</author>
    <category>english</category><category>3d-gaussian-splatting</category><category>photogrammetry</category><category>nerfstudio</category><category>postshot</category><category>luma-ai</category><category>polycam</category><category>apple-object-capture</category><category>realitycapture</category><category>3d-reconstruction</category>
  </item>

  <item>
    <guid>https://www.youngju.dev/blog/culture/2026-05-16-3d-gaussian-splatting-photogrammetry-2026-nerfstudio-postshot-luma-genie-polycam-apple-object-capture-realitycapture-deep-dive.ja</guid>
    <title>3Dガウシアン・スプラッティングとフォトグラメトリ 2026 完全ガイド - Nerfstudio · Postshot · Luma Genie · Polycam · Apple Object Capture · RealityCapture 徹底解説</title>
    <link>https://www.youngju.dev/blog/culture/2026-05-16-3d-gaussian-splatting-photogrammetry-2026-nerfstudio-postshot-luma-genie-polycam-apple-object-capture-realitycapture-deep-dive.ja</link>
    <description>2023年のINRIA論文がNeRFの時代を閉じた。2026年には、写真100枚と一台のデスクトップで誰でも写真のような3Dを焼ける。Nerfstudio 1.4のSplatfacto、Postshot 2.0、Luma Genie 2、Polycam、macOS SequoiaのObject Capture API、無料化したRealityCapture 1.5、Scaniverse 4、Meshroom。キャプチャからWebビューワまでを一息で整理する。</description>
    <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
    <author>fjvbn2003@gmail.com (Youngju Kim)</author>
    <category>日本語</category><category>3d-gaussian-splatting</category><category>photogrammetry</category><category>nerfstudio</category><category>postshot</category><category>luma-ai</category><category>polycam</category><category>apple-object-capture</category><category>realitycapture</category><category>3d-reconstruction</category>
  </item>

  <item>
    <guid>https://www.youngju.dev/blog/culture/2026-05-16-3d-gaussian-splatting-photogrammetry-2026-nerfstudio-postshot-luma-genie-polycam-apple-object-capture-realitycapture-deep-dive</guid>
    <title>3D 가우시안 스플래팅 &amp; 포토그래메트리 2026 완벽 가이드 - Nerfstudio · Postshot · Luma Genie · Polycam · Apple Object Capture · RealityCapture 심층 분석</title>
    <link>https://www.youngju.dev/blog/culture/2026-05-16-3d-gaussian-splatting-photogrammetry-2026-nerfstudio-postshot-luma-genie-polycam-apple-object-capture-realitycapture-deep-dive</link>
    <description>2023년 INRIA의 3DGS 논문이 NeRF의 시대를 닫았다. 2026년에는 사진 100장이면 누구나 데스크톱에서 사진실적 3D를 굽는다. Nerfstudio 1.x의 Splatfacto, Postshot의 .ksplat, Luma Genie의 텍스트-투-3D, Polycam의 LiDAR 캡처, Apple Object Capture의 macOS Sequoia API, 무료가 된 RealityCapture 1.5, Scaniverse 4, Meshroom — 캡처에서 웹 뷰어까지 한 호흡으로 정리한다.</description>
    <pubDate>Sat, 16 May 2026 00:00:00 GMT</pubDate>
    <author>fjvbn2003@gmail.com (Youngju Kim)</author>
    <category>3d-gaussian-splatting</category><category>photogrammetry</category><category>nerfstudio</category><category>postshot</category><category>luma-ai</category><category>polycam</category><category>apple-object-capture</category><category>realitycapture</category><category>3d-reconstruction</category>
  </item>

    </channel>
  </rss>
