{
  "docId": "019dd923-5e88-73ef-bd5c-fd5012384ae3",
  "docSlug": "bf350dd574c19997",
  "documentTitle": "2021 Air Street Capital The State of AI Report 2021",
  "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": 40,
  "pageCount": 188,
  "prevPage": 39,
  "nextPage": 41,
  "slideType": "case_study",
  "function": "illustrate_case",
  "density": "balanced",
  "nDataPoints": 0,
  "notes": "The slide details the VLiD architecture, showing training and inference workflows.",
  "elementsJson": [
    "bullet_list",
    "flowchart",
    "callout_box"
  ],
  "metadataConfidence": 1,
  "imagePath": null,
  "slideHref": "/slides/019dd923-5e88-73ef-bd5c-fd5012384ae3/40",
  "deckHref": "/decks/019dd923-5e88-73ef-bd5c-fd5012384ae3",
  "deckJsonHref": "/decks/019dd923-5e88-73ef-bd5c-fd5012384ae3.json",
  "deckAnchorHref": "/decks/019dd923-5e88-73ef-bd5c-fd5012384ae3#slide-40",
  "components": [
    {
      "bbox": null,
      "kind": "callout",
      "text": "CLIP is already serving as a base model for downstream tasks: Researchers from Google use its zero-shot capabilities together with Mask R-CNN to create a zero-shot learning model (VLiD) that surpasses supervised models on zero-shot object detection.",
      "attrs": null,
      "subkind": null,
      "toolName": "Visual emphasis",
      "toolSlug": "visual-emphasis",
      "confidence": null,
      "componentId": "019dd952-47c3-7407-8d26-f70a677563b7",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.45,
        "w": 0.48,
        "x": 0.5,
        "y": 0.38
      },
      "kind": "image",
      "text": "Diagram showing Training and Inference workflows for VLiD",
      "attrs": null,
      "subkind": "infographic",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "fecf72cc-4ce9-4bfe-ae42-4c00e494e247",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.35,
        "w": 0.45,
        "x": 0.02,
        "y": 0.33
      },
      "kind": "list",
      "text": "During training, VLiD is only given a part of the classes to predict, for which CLIP generates class representations. Then, VLiD uses CLIP to predict the class of the image representations generated by Mask R-CNN.\nOnly during inference is VLiD given the novel classes that were unseen during training.\nVLiD is the first zero-shot object detector to be evaluated on the LVIS dataset, and it outperformed its supervised counterpart on the novel categories.",
      "attrs": null,
      "subkind": "bullet",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "0be715bd-3bfe-4622-ad58-f066fba7f9b8",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.09,
        "w": 0.95,
        "x": 0.02,
        "y": 0.21
      },
      "kind": "paragraph",
      "text": "CLIP is already serving as a base model for downstream tasks: Researchers from Google use its zero-shot capabilities together with Mask R-CNN to create a zero-shot learning model (VLiD) that surpasses supervised models on zero-shot object detection.",
      "attrs": null,
      "subkind": "paragraph",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "40951503-b775-406e-afb7-66daf0339fec",
      "frameworkName": null,
      "frameworkSlug": null
    },
    {
      "bbox": {
        "h": 0.04,
        "w": 0.6,
        "x": 0.02,
        "y": 0.14
      },
      "kind": "title",
      "text": "Using CLIP’s learned representations for zero-shot object detection",
      "attrs": null,
      "subkind": "headline",
      "toolName": null,
      "toolSlug": null,
      "confidence": null,
      "componentId": "fb48f6e4-7135-435d-8ead-080fd98b985e",
      "frameworkName": null,
      "frameworkSlug": null
    }
  ],
  "metrics": [],
  "tools": [],
  "frameworks": [],
  "arcBeats": [
    {
      "to": 152,
      "from": 5,
      "beatId": "019dd95a-0682-776c-8e35-0affbadec38e",
      "arcName": "The Triple Take",
      "arcSlug": "triple-take",
      "beatName": "The Facts (What)",
      "beatSlug": "triple-take-the-facts-what",
      "evidence": "Definitions p.5-6, Exec Summary p.7, then Sections 1-3 catalog evidence.",
      "position": 1,
      "confidence": 78,
      "parentBeatName": "Setup",
      "parentBeatSlug": "setup"
    },
    {
      "to": 152,
      "from": 10,
      "beatId": "019dd95a-0682-776c-8e35-1b3ee3a1c012",
      "arcName": "The Mountain",
      "arcSlug": "mountain",
      "beatName": "Rising Action",
      "beatSlug": "mountain-rising-action",
      "evidence": "Three sections accumulating evidence of accelerating progress.",
      "position": 2,
      "confidence": 55,
      "parentBeatName": "Development",
      "parentBeatSlug": "development"
    }
  ],
  "loops": [
    {
      "to": 40,
      "from": 36,
      "name": "Aha Moment",
      "slug": "03-aha-moment",
      "bestFor": "Data-heavy sections, research findings, analytical arguments",
      "matchId": "019dd95a-07fe-70ce-8d3d-53192fc75199",
      "evidence": "Diffusion models > GANs, then CLIP, then DALL-E payoff.",
      "position": 5,
      "objective": "Reveal multimodal generative AI breakthrough",
      "structure": "The Problem/Question -> What the Data Shows -> The Insight",
      "confidence": 72,
      "description": "Establish a tension, present data, then deliver the insight that changes everything"
    }
  ],
  "imagePathAlt": null,
  "thumbSrc": null,
  "thumbSrcAlt": null,
  "locked": true
}