Hugging Face's logo

qualcomm
/
QuickSRNetSmall

Image-to-Image
PyTorch
android
Model card Files
xet
 Community
QuickSRNetSmall
File size: 13,123 Bytes
79f470b
 
 
 
 
f8da907
79f470b
 
 
a71ce34
79f470b
c95eb37
e815d88
79f470b
 
c95eb37
fe6cfc3
c95eb37
 
 
 
 
 
 
 
 
 
 
 
fe6cfc3
 
 
 
 
 
c95eb37
 
 
 
 
 
fe6cfc3
c95eb37
 
 
 
 
 
fe6cfc3
c95eb37
 
 
 
 
 
 
 
 
 
 
 
 
 
 
fe6cfc3
 
 
 
 
 
 
 
c3dc8f5
fe6cfc3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8e8123f
79f470b
3bcc01a
9c155b2
8e8123f
79f470b
 
 
 
 
83eb19d
79f470b
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
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
---
library_name: pytorch
license: other
tags:
- android
pipeline_tag: image-to-image

---

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

# QuickSRNetSmall: Optimized for Qualcomm Devices

QuickSRNet Small is designed for upscaling images on mobile platforms to sharpen in real-time.

This is based on the implementation of QuickSRNetSmall found [here](https://github.com/quic/aimet-model-zoo/tree/develop/aimet_zoo_torch/quicksrnet).
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/v0.59.0/src/qai_hub_models/models/quicksrnetsmall) 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 |
|---|---|---|---|---|
| ONNX | float | Universal | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/quicksrnetsmall/releases/v0.59.0/quicksrnetsmall-onnx-float.zip)
| ONNX | w8a8 | Universal | QAIRT 2.45, ONNX Runtime 1.27.1 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/quicksrnetsmall/releases/v0.59.0/quicksrnetsmall-onnx-w8a8.zip)
| QNN_DLC | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/quicksrnetsmall/releases/v0.59.0/quicksrnetsmall-qnn_dlc-float.zip)
| QNN_DLC | w8a8 | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/quicksrnetsmall/releases/v0.59.0/quicksrnetsmall-qnn_dlc-w8a8.zip)
| TFLITE | float | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/quicksrnetsmall/releases/v0.59.0/quicksrnetsmall-tflite-float.zip)
| TFLITE | w8a8 | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/quicksrnetsmall/releases/v0.59.0/quicksrnetsmall-tflite-w8a8.zip)

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


### Option 2: Export with Custom Configurations

Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/v0.59.0/src/qai_hub_models/models/quicksrnetsmall) 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 [QuickSRNetSmall on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.59.0/src/qai_hub_models/models/quicksrnetsmall) for usage instructions.

## Model Details

**Model Type:** Model_use_case.super_resolution

**Model Stats:**
- Model checkpoint: quicksrnet_small_3x_checkpoint
- Input resolution: 128x128
- Number of parameters: 33.3K
- Model size (float): 133 KB
- Model size (w8a8): 41.7 KB

## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| QuickSRNetSmall | ONNX | float | Snapdragon® X2 Elite | 0.361 ms | 8 - 8 MB | NPU
| QuickSRNetSmall | ONNX | float | Snapdragon® X Elite | 0.86 ms | 8 - 8 MB | NPU
| QuickSRNetSmall | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 0.543 ms | 0 - 28 MB | NPU
| QuickSRNetSmall | ONNX | float | Snapdragon® 8 Gen 1 Mobile | 1.222 ms | 0 - 35 MB | NPU
| QuickSRNetSmall | ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 0.851 ms | 0 - 2 MB | NPU
| QuickSRNetSmall | ONNX | float | Qualcomm® QCS8450 | 1.222 ms | 0 - 35 MB | NPU
| QuickSRNetSmall | ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 1.107 ms | 6 - 9 MB | NPU
| QuickSRNetSmall | ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 0.86 ms | 8 - 8 MB | NPU
| QuickSRNetSmall | ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 0.434 ms | 0 - 17 MB | NPU
| QuickSRNetSmall | ONNX | float | Snapdragon® 8 Elite Mobile | 0.434 ms | 0 - 17 MB | NPU
| QuickSRNetSmall | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.39 ms | 0 - 18 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Snapdragon® X2 Elite | 0.191 ms | 3 - 3 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Snapdragon® X Elite | 0.486 ms | 3 - 3 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.314 ms | 0 - 29 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Snapdragon® 8 Gen 1 Mobile | 0.568 ms | 0 - 34 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Qualcomm® Dragonwing™ QCS6490 | 2.103 ms | 0 - 3 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 0.502 ms | 0 - 2 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Qualcomm® QCS8450 | 0.568 ms | 0 - 34 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Qualcomm® Dragonwing™ IQ-9075 | 0.643 ms | 0 - 3 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Qualcomm® Dragonwing™ IQ-X7181 | 0.486 ms | 3 - 3 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Qualcomm® Dragonwing™ Q-8750 | 0.238 ms | 0 - 22 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Snapdragon® 8 Elite Mobile | 0.238 ms | 0 - 22 MB | NPU
| QuickSRNetSmall | ONNX | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.214 ms | 0 - 18 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Snapdragon® X2 Elite | 0.407 ms | 0 - 0 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Snapdragon® X Elite | 0.829 ms | 0 - 0 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 0.445 ms | 0 - 27 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Snapdragon® 8 Gen 1 Mobile | 1.041 ms | 0 - 34 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Qualcomm® Dragonwing™ QCS8275 | 1.87 ms | 0 - 20 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 0.744 ms | 0 - 53 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Qualcomm® SA8775P | 1.071 ms | 0 - 22 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Qualcomm® SA8650P | 1.071 ms | 0 - 22 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Qualcomm® SA8255P | 1.071 ms | 0 - 22 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Qualcomm® QCS8450 | 1.041 ms | 0 - 34 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-9075 | 0.961 ms | 0 - 5 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-X7181 | 0.829 ms | 0 - 0 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Qualcomm® Dragonwing™ Q-8750 | 0.354 ms | 0 - 20 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Qualcomm® SA7255P | 1.87 ms | 0 - 20 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Qualcomm® SA8295P | 1.38 ms | 0 - 17 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 0.354 ms | 0 - 20 MB | NPU
| QuickSRNetSmall | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.336 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Snapdragon® X2 Elite | 0.246 ms | 0 - 0 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Snapdragon® X Elite | 0.445 ms | 0 - 0 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.224 ms | 0 - 27 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Snapdragon® 8 Gen 1 Mobile | 0.42 ms | 0 - 32 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ QCS6490 | 1.05 ms | 0 - 2 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ QCS8275 | 0.835 ms | 0 - 19 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 0.356 ms | 0 - 8 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® SA8775P | 0.542 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® SA8650P | 0.542 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® SA8255P | 0.542 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® QCS8450 | 0.42 ms | 0 - 32 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ IQ-9075 | 0.474 ms | 0 - 2 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ IQ-X7181 | 0.445 ms | 0 - 0 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ Q-6690 | 1.276 ms | 0 - 16 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ Q-7790 | 0.371 ms | 0 - 18 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ Q-8750 | 0.195 ms | 0 - 20 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® SA7255P | 0.835 ms | 0 - 19 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Qualcomm® SA8295P | 0.754 ms | 0 - 17 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Snapdragon® 8 Elite Mobile | 0.195 ms | 0 - 20 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.165 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.371 ms | 0 - 18 MB | NPU
| QuickSRNetSmall | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 0.613 ms | 0 - 28 MB | NPU
| QuickSRNetSmall | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 1.309 ms | 0 - 33 MB | NPU
| QuickSRNetSmall | TFLITE | float | Qualcomm® Dragonwing™ QCS8275 | 2.362 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.0 ms | 0 - 3 MB | NPU
| QuickSRNetSmall | TFLITE | float | Qualcomm® SA8775P | 1.452 ms | 0 - 22 MB | NPU
| QuickSRNetSmall | TFLITE | float | Qualcomm® SA8650P | 1.452 ms | 0 - 22 MB | NPU
| QuickSRNetSmall | TFLITE | float | Qualcomm® SA8255P | 1.452 ms | 0 - 22 MB | NPU
| QuickSRNetSmall | TFLITE | float | Qualcomm® QCS8450 | 1.309 ms | 0 - 33 MB | NPU
| QuickSRNetSmall | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 1.49 ms | 1 - 6 MB | NPU
| QuickSRNetSmall | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 0.462 ms | 0 - 20 MB | NPU
| QuickSRNetSmall | TFLITE | float | Qualcomm® SA7255P | 2.362 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | TFLITE | float | Qualcomm® SA8295P | 1.643 ms | 0 - 17 MB | NPU
| QuickSRNetSmall | TFLITE | float | Snapdragon® 8 Elite Mobile | 0.462 ms | 0 - 20 MB | NPU
| QuickSRNetSmall | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.395 ms | 0 - 22 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.268 ms | 0 - 27 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Snapdragon® 8 Gen 1 Mobile | 0.471 ms | 0 - 33 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® Dragonwing™ QCS6490 | 0.936 ms | 0 - 2 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® Dragonwing™ QCS8275 | 0.877 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 0.403 ms | 0 - 1 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® SA8775P | 0.595 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® SA8650P | 0.595 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® SA8255P | 0.595 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® QCS8450 | 0.471 ms | 0 - 33 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® Dragonwing™ IQ-9075 | 0.551 ms | 0 - 3 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® Dragonwing™ Q-6690 | 1.507 ms | 0 - 16 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® Dragonwing™ Q-7790 | 0.419 ms | 0 - 18 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® Dragonwing™ Q-8750 | 0.225 ms | 0 - 19 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® SA7255P | 0.877 ms | 0 - 21 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Qualcomm® SA8295P | 0.793 ms | 0 - 17 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Snapdragon® 8 Elite Mobile | 0.225 ms | 0 - 19 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.175 ms | 0 - 24 MB | NPU
| QuickSRNetSmall | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.419 ms | 0 - 18 MB | NPU

## License
* The license for the original implementation of QuickSRNetSmall can be found
  [here](https://github.com/quic/aimet-model-zoo/blob/develop/LICENSE.md).

## References
* [QuickSRNet: Plain Single-Image Super-Resolution Architecture for Faster Inference on Mobile Platforms](https://arxiv.org/abs/2303.04336)
* [Source Model Implementation](https://github.com/quic/aimet-model-zoo/tree/develop/aimet_zoo_torch/quicksrnet)

## 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).