Hugging Face's logo

qualcomm
/
OWL-V2

Object Detection
PyTorch
foundation
android
Model card Files
xet
 Community
OWL-V2
File size: 14,648 Bytes
6c0bdc5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
---
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).