NetraFlow
Capabilities

Brands

Detect brand appearances in video with confidence scores, timestamps, and proof frames.

Brand detection identifies brands appearing visually on screen or mentioned in speech. Both video frames and audio are analyzed to find brand appearances with confidence scores and timestamps.

curl -X POST https://api.netraflow.com/v1/jobs \
  -H "Content-Type: application/json" \
  -H "X-Api-Key: sk_live_your_key_here" \
  -d '{
    "url": "https://www.youtube.com/watch?v=example",
    "capabilities": ["brands"]
  }'
{
  "data": {
    "job_id": "job_abc123",
    "status": "completed",
    "results": {
      "brands": [
        {
          "name": "Nike",
          "confidence": 0.95,
          "total_appearances": 3,
          "total_visible_duration": 8.4,
          "first_seen": 12.5,
          "last_seen": 145.0,
          "appearances": [
            {
              "timestamp": 12.5,
              "timestamp_start": 12.0,
              "timestamp_end": 15.5,
              "visible_duration": 3.5,
              "context": "Nike logo on athlete's jersey",
              "source": "visual"
            },
            {
              "timestamp": 89.0,
              "context": "Speaker mentions Nike partnership",
              "speech_context": "We partnered with Nike for this campaign",
              "source": "audio"
            },
            {
              "timestamp": 145.0,
              "timestamp_start": 143.0,
              "timestamp_end": 148.0,
              "visible_duration": 5.0,
              "context": "Nike swoosh on product close-up",
              "source": "visual"
            }
          ]
        },
        {
          "name": "Adidas",
          "confidence": 0.72,
          "total_appearances": 1,
          "first_seen": 67.0,
          "last_seen": 67.0,
          "appearances": [
            {
              "timestamp": 67.0,
              "context": "Adidas shoe visible in background",
              "source": "visual"
            }
          ]
        }
      ]
    }
  }
}

Brand fields

Prop

Type

Appearance fields

Prop

Type

Proof frames

Set return_proof: true in your request to receive proof frame URLs for high-confidence detections. Frames are generated for appearances with confidence >= 0.8, verified by a secondary AI pass, with up to 20 frames per job.

curl -X POST https://api.netraflow.com/v1/jobs \
  -H "Content-Type: application/json" \
  -H "X-Api-Key: sk_live_your_key_here" \
  -d '{
    "url": "https://www.youtube.com/watch?v=example",
    "capabilities": ["brands"],
    "return_proof": true
  }'
{
  "data": {
    "job_id": "job_abc123",
    "status": "completed",
    "results": {
      "brands": [
        {
          "name": "Nike",
          "confidence": 0.95,
          "total_appearances": 2,
          "appearances": [
            {
              "timestamp": 12.5,
              "frame_url": "/v1/frames/job_abc123/brand_nike_12.5.jpg",
              "bounding_box": { "x": 0.3, "y": 0.2, "width": 0.15, "height": 0.1 },
              "context": "Nike logo on jersey",
              "source": "visual"
            },
            {
              "timestamp": 145.0,
              "frame_url": "/v1/frames/job_abc123/brand_nike_145.0.jpg",
              "context": "Nike swoosh on product",
              "source": "visual"
            }
          ]
        }
      ]
    }
  }
}

Proof frame URLs are authenticated — fetch them with the same X-Api-Key header.

On this page