---
library_name: pytorch
license: apache-2.0
tags:
- foundation
- android
pipeline_tag: object-detection

---

![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/web-assets/model_demo.png)

# OWL-V2: Optimized for Qualcomm Devices

OWL-V2 (Open-World Localization with Vision Transformers) is an open-vocabulary object detector that uses a CLIP-based ViT-B/16 backbone. Given an image and one or more free-form text queries, the model predicts bounding boxes and confidence scores for each query.

This is based on the implementation of OWL-V2 found [here](https://github.com/google-research/scenic/tree/main/scenic/projects/owl_vit).
This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.61.0/src/qai_hub_models/models/owlv2) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).

Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) to run these models on a hosted Qualcomm® device.

## Getting Started
There are two ways to deploy this model on your device:

### Option 1: Download Pre-Exported Models

Below are pre-exported model assets ready for deployment.

| Runtime | Precision | Chipset | SDK Versions | Download |
|---|---|---|---|---|
| PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-precompiled_qnn_onnx-float-qualcomm_snapdragon_x2_elite.zip)
| PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-precompiled_qnn_onnx-float-qualcomm_snapdragon_x_elite.zip)
| PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-precompiled_qnn_onnx-float-qualcomm_snapdragon_8gen3.zip)
| PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 1 Mobile | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-precompiled_qnn_onnx-float-qualcomm_snapdragon_8gen1.zip)
| PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ IQ-8275 | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-precompiled_qnn_onnx-float-qualcomm_qcs8275.zip)
| PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-precompiled_qnn_onnx-float-qualcomm_qcs8550_proxy.zip)
| PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-precompiled_qnn_onnx-float-qualcomm_qcs9075.zip)
| PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Mobile | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-precompiled_qnn_onnx-float-qualcomm_snapdragon_8_elite_for_galaxy.zip)
| PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-precompiled_qnn_onnx-float-qualcomm_snapdragon_8_elite_gen5_for_galaxy.zip)
| QNN_CONTEXT_BINARY | float | Snapdragon® X2 Elite | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_snapdragon_x2_elite.zip)
| QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_snapdragon_x_elite.zip)
| QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_snapdragon_8gen3.zip)
| QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 1 Mobile | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_snapdragon_8gen1.zip)
| QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ IQ-8275 | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_qcs8275.zip)
| QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_qcs8550_proxy.zip)
| QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_sa8775p.zip)
| QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ IQ-9075 | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_qcs9075.zip)
| QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_sa7255p.zip)
| QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_sa8295p.zip)
| QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Mobile | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_snapdragon_8_elite_for_galaxy.zip)
| QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/owlv2/releases/v0.61.0/owlv2-qnn_context_binary-float-qualcomm_snapdragon_8_elite_gen5_for_galaxy.zip)

For more device-specific assets and performance metrics, visit **[OWL-V2 on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/owlv2)**.


### Option 2: Export with Custom Configurations

Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.61.0/src/qai_hub_models/models/owlv2) Python library to compile and export the model with your own:
- Custom weights (e.g., fine-tuned checkpoints)
- Custom input shapes
- Target device and runtime configurations

This option is ideal if you need to customize the model beyond the default configuration provided here.

See our repository for [OWL-V2 on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.61.0/src/qai_hub_models/models/owlv2) for usage instructions.

## Model Details

**Model Type:** Model_use_case.object_detection

**Model Stats:**
- Input image resolution: 960x960
- Model checkpoint: google/owlv2-base-patch16-ensemble
- Model size (float): 620 MB
- Number of parameters: 155M

## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| text | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 3.281 ms | 11 - 11 MB | NPU
| text | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 5.983 ms | 116 - 116 MB | NPU
| text | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 4.109 ms | 11 - 23 MB | NPU
| text | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 1 Mobile | 13.253 ms | 11 - 25 MB | NPU
| text | PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ IQ-8275 | 8.026 ms | 10 - 24 MB | NPU
| text | PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 5.942 ms | 11 - 12 MB | NPU
| text | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8450 | 13.253 ms | 11 - 25 MB | NPU
| text | PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 7.111 ms | 11 - 24 MB | NPU
| text | PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 5.983 ms | 116 - 116 MB | NPU
| text | PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 3.661 ms | 2 - 15 MB | NPU
| text | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Mobile | 3.661 ms | 2 - 15 MB | NPU
| text | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.407 ms | 0 - 7 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Snapdragon® X2 Elite | 3.782 ms | 11 - 11 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 6.188 ms | 11 - 11 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 4.106 ms | 11 - 18 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 1 Mobile | 13.185 ms | 9 - 19 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ IQ-8275 | 7.89 ms | 11 - 24 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ IQ-8275 | 18.899 ms | 1 - 10 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 5.941 ms | 11 - 12 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | 7.329 ms | 0 - 9 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® SA8650P | 7.329 ms | 0 - 9 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® SA8255P | 7.329 ms | 0 - 9 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8450 | 13.185 ms | 9 - 19 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ IQ-9075 | 7.035 ms | 11 - 23 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ IQ-X7181 | 6.188 ms | 11 - 11 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ Q-8750 | 3.648 ms | 0 - 9 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | 18.899 ms | 1 - 10 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | 12.423 ms | 3 - 8 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Mobile | 3.648 ms | 0 - 9 MB | NPU
| text | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 3.401 ms | 10 - 19 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 626.064 ms | 33 - 33 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 1223.664 ms | 295 - 295 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 979.372 ms | 11 - 23 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 1 Mobile | 2158.25 ms | 17 - 31 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ IQ-8275 | 1229.848 ms | 11 - 25 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1399.478 ms | 0 - 310 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8450 | 2158.25 ms | 17 - 31 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 1208.655 ms | 10 - 24 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 1223.664 ms | 295 - 295 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 757.463 ms | 13 - 25 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Mobile | 757.463 ms | 13 - 25 MB | NPU
| vision | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 741.23 ms | 13 - 26 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Snapdragon® X2 Elite | 600.765 ms | 11 - 11 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 1246.656 ms | 11 - 11 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 965.04 ms | 11 - 19 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 1 Mobile | 2119.442 ms | 11 - 25 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ IQ-8275 | 1230.158 ms | 11 - 34 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ IQ-8275 | 2456.487 ms | 1 - 10 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1428.974 ms | 11 - 14 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | 1266.806 ms | 0 - 10 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® SA8650P | 1266.806 ms | 0 - 10 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® SA8255P | 1266.806 ms | 0 - 10 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8450 | 2119.442 ms | 11 - 25 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ IQ-9075 | 1208.734 ms | 13 - 35 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ IQ-X7181 | 1246.656 ms | 11 - 11 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® Dragonwing™ Q-8750 | 699.47 ms | 11 - 21 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | 2456.487 ms | 1 - 10 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | 1421.1 ms | 0 - 10 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Mobile | 699.47 ms | 11 - 21 MB | NPU
| vision | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 628.172 ms | 11 - 19 MB | NPU

## License
* The license for the original implementation of OWL-V2 can be found
  [here](https://github.com/google-research/scenic/blob/main/LICENSE).

## References
* [Scaling Open-Vocabulary Object Detection](https://arxiv.org/abs/2306.09683)
* [Source Model Implementation](https://github.com/google-research/scenic/tree/main/scenic/projects/owl_vit)

## Community
* Join [our AI Hub Slack community](https://aihub.qualcomm.com/community/slack) to collaborate, post questions and learn more about on-device AI.
* For questions or feedback please [reach out to us](mailto:ai-hub-support@qti.qualcomm.com).