{
  "docId": "019dd923-5e88-73ef-bd5d-06b04d219fea",
  "docSlug": "dd91c78f6570bf29",
  "documentTitle": "2023 Air Street Capital The State of AI Report 2023",
  "authorId": "AirStreetCapital",
  "authorName": "Air Street Capital",
  "documentKindSlug": "consulting-deck",
  "documentKindLabel": "Consulting deck",
  "sourceTypeSlug": "vc_research",
  "sourceTypeLabel": "VC research",
  "presentationDate": null,
  "orientation": "landscape",
  "aspectRatio": 1.777,
  "pageNumber": 56,
  "pageCount": 163,
  "prevPage": 55,
  "nextPage": 57,
  "slideType": "industry_trends",
  "function": "summarize",
  "density": "dense",
  "nDataPoints": 2,
  "notes": "The slide highlights technical advancements in self-supervised learning, specifically dataset curation and model stability.",
  "elementsJson": [
    "headline_text",
    "bullet_list",
    "photo"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5d-06b04d219fea/56",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5d-06b04d219fea",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5d-06b04d219fea.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5d-06b04d219fea#slide-56",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "DINOv2 is a self-supervised Vision Transformer model from Meta, producing universal visual features that can be used across a variety of image level and pixel level tasks without fine-tuning.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c4-7719-a225-72608e68541a",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.25,
        "w": 0.35,
        "x": 0.625,
        "y": 0.5
      },
      "kind": "image",
      "text": "Comparison of DINO vs DINOv2 feature maps on a dog image.",
      "attrs": null,
      "subkind": "illustration",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "7765da46-e236-43be-922d-1955e82e87bf",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.7,
        "w": 0.6,
        "x": 0.02,
        "y": 0.2
      },
      "kind": "list",
      "text": "DINOv2 is a self-supervised Vision Transformer model from Meta, producing universal visual features that can be used across a variety of image level (e.g. classification) and pixel level (e.g. segmentation) tasks without fine-tuning and are competitive with SOTA open-source weakly supervised alternatives.\nIt is the first work to close the gap between self-supervised and weakly supervised approaches. DINOv2 features are shown to contain information about object parts as well as semantic and low level understanding of images.\nThe authors made the training of self-supervised learning models more stable through additional regularisation methods and reduced the memory requirements, which enabled training larger models on more data for longer. They also provide compressed versions of the models obtained through distillation.\nAlthough any image can be used for training, a key component was curating the dataset and automatically balancing it across concepts (keeping 142M out of 1.2B source images). Re-visit this slide.\nDINOv2 features can be used with linear classifiers to obtain strong results across many visual tasks.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "b135c92f-bb3c-4864-8b1f-e9d030df80d5",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.04,
        "w": 0.5,
        "x": 0.02,
        "y": 0.14
      },
      "kind": "title",
      "text": "DINOv2: the new default computer vision backbone",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "904a9bed-0158-40a3-963b-46e49a630f21",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [
    {
      "name": "Chartjunk Elimination",
      "slug": "chartjunk-elimination",
      "agent": "Designer",
      "layer": "slide",
      "matchId": "36e0d60d-d069-4008-af67-37bf4549281a",
      "evidence": "Comparison of DINO vs DINOv2 feature maps on a dog image.",
      "confidence": 0.7
    }
  ],
  "frameworks": [],
  "arcBeats": [
    {
      "to": 120,
      "from": 11,
      "beatId": "019dd95a-0682-776c-8e35-41afd44ef59f",
      "arcName": "The Triple Take",
      "arcSlug": "triple-take",
      "beatName": "The Facts (What)",
      "beatSlug": "triple-take-the-facts-what",
      "evidence": "Research + Industry sections inventory model, compute, funding facts.",
      "position": 1,
      "confidence": 78,
      "parentBeatName": "Setup",
      "parentBeatSlug": "setup"
    },
    {
      "to": 120,
      "from": 11,
      "beatId": "019dd95a-0682-776c-8e35-523bfb7f96e6",
      "arcName": "The Mountain",
      "arcSlug": "mountain",
      "beatName": "Rising Action",
      "beatSlug": "mountain-rising-action",
      "evidence": "Escalating capabilities, compute concentration and capital flows.",
      "position": 2,
      "confidence": 45,
      "parentBeatName": "Development",
      "parentBeatSlug": "development"
    }
  ],
  "loops": [
    {
      "to": 58,
      "from": 50,
      "name": "Pattern Hunter",
      "slug": "02-pattern-hunter",
      "bestFor": "Time-pressed audiences, building consensus, when data is strong",
      "matchId": "019dd95a-07fe-70ce-8d3d-f70f6435b722",
      "evidence": "Text-to-video, instruction editing, Gaussian splatting, NeRFs, depth, SAM, DINOv2, weather, music.",
      "position": 6,
      "objective": "Generative media and visual foundation models proliferate",
      "structure": "Evidence A -> Evidence B -> Evidence C -> Pattern/Conclusion",
      "confidence": 72,
      "description": "Group multiple pieces of evidence that together point to a pattern or conclusion"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}