detr-resnet-50

Beta

Model ID: @cf/facebook/detr-resnet-50

DEtection TRansformer (DETR) model trained end-to-end on COCO 2017 object detection (118k annotated images).

Properties

Task Type: Object Detection

Code Examples

Workers - Typescript

export interface Env {
  AI: Ai;
}

export default {
  async fetch(request, env): Promise<Response> {
    const res = await fetch("https://cataas.com/cat");
    const blob = await res.arrayBuffer();

    const inputs = {
      image: [...new Uint8Array(blob)],
    };

    const response = await env.AI.run(
      "@cf/facebook/detr-resnet-50",
      inputs
    );

    return new Response(JSON.stringify({ inputs: { image: [] }, response }));
  },
} satisfies ExportedHandler<Env>;

curl

curl https://api.cloudflare.com/client/v4/accounts/$CLOUDFLARE_ACCOUNT_ID/ai/run/@cf/facebook/detr-resnet-50  \
    -X POST  \
    -H "Authorization: Bearer $CLOUDFLARE_API_TOKEN"  \
    --data-binary "@pedestrian-boulevard-manhattan-crossing.jpg"

Response

This task returns a list of detected objects, each one containing a label, a probability score, and its surrounding box coordinates.

[
  { "label":"cat" ,"score":0.4071170687675476, "box": { "xmin": 0, "ymin": 0, "xmax": 10, "ymax": 10 } },
  { "label":"face", "score":0.22562485933303833, "box": { "xmin": 15, "ymin": 22, "xmax": 25, "ymax": 35 } },
  { "label":"car", "score":0.033316344022750854, "box": { "xmin": 72, "ymin": 55, "xmax": 95, "ymax": 72 } }
]

API Schema

The following schema is based on JSON Schema

Input JSON Schema

{
  "oneOf": [
    {
      "type": "string",
      "format": "binary"
    },
    {
      "type": "object",
      "properties": {
        "image": {
          "type": "array",
          "items": {
            "type": "number"
          }
        }
      }
    }
  ]
}

Output JSON Schema

{
  "type": "array",
  "contentType": "application/json",
  "items": {
    "type": "object",
    "properties": {
      "score": {
        "type": "number"
      },
      "label": {
        "type": "string"
      },
      "box": {
        "type": "object",
        "properties": {
          "xmin": {
            "type": "number"
          },
          "ymin": {
            "type": "number"
          },
          "xmax": {
            "type": "number"
          },
          "ymax": {
            "type": "number"
          }
        }
      }
    }
  }
}