owlv2-large-patch14-ensemble

Name: owlv2-large-patch14-ensemble
Author: google

by google

Open-vocabulary OWLv2 detector for text-conditioned object search

145Kdl/month

46likes

438Mparams

HuggingFace Run it on your own data, free

Identifiers

Model ID

google/owlv2-large-patch14-ensemble

Feature URI

mixpeek://image_extractor@v1/google_owlv2_large_ensemble_v1

Overview

OWLv2 Large Patch14 Ensemble is Google's open-vocabulary detector for zero-shot object localization. It lets a pipeline search for objects described in text instead of relying only on a fixed supervised label set.

On Mixpeek, OWLv2 is useful when an agent needs to find visual categories that change by task: a specific product shape, a UI control, damaged equipment, or a visual policy violation. The detector outputs boxes and labels that can be stored, filtered, and joined with embeddings or captions.

Architecture

Vision Transformer based open-vocabulary object detector. It aligns text queries and image regions so arbitrary text labels can guide detection at inference time.

Mixpeek SDK Integration

// No extractor parameter takes a Hugging Face model id (checked against
// GET /v1/discovery/extractors, which returns 13), so owlv2-large-patch14-ensemble runs
// on your side and the output is upserted through POST
// /v1/namespaces/{namespace_id}/documents/upsert. On Enterprise the other
// path is to upload the weights instead: POST /v1/namespaces/{id}/models
// accepts the huggingface format and a custom plugin loads them.
const res = await fetch(
  "https://api.mixpeek.com/v1/namespaces/ns_your_namespace/documents/upsert",
  {
    method: "POST",
    headers: {
      Authorization: "Bearer API_KEY",
      "Content-Type": "application/json",
    },
    body: JSON.stringify({
      collection_id: "col_your_collection",
      documents: [
        {
          document_id: "asset-00412",
          // Boxes, masks, depth maps and anomaly scores are structured
          // results, not vectors. They go in payload and are reachable
          // through pre_filters on a retriever, not through similarity.
          payload: {
            detections: modelOutput,
            source_key: "archive/2026/asset-00412",
          },
        },
      ],
    }),
  },
);

// No managed alternative for an open label set. Two extractors do emit a
// bbox, for the one thing each detects: document_graph_extractor@v1 per
// layout block, face_identity_extractor@v1 per face. Nothing ships that
// returns masks, depth maps or anomaly scores.