Hugging Face's logo

qualcomm
/
QuickSRNetLarge

Image-to-Image
PyTorch
android
Model card Files
xet
 Community
QuickSRNetLarge
File size: 13,132 Bytes
fe32929
 
 
 
 
3b79ef9
fe32929
 
 
9f647b5
fe32929
cdd46c4
950b6b2
fe32929
 
cdd46c4
4ae50d3
cdd46c4
 
 
 
 
 
 
 
 
 
 
 
4ae50d3
 
 
 
 
 
cdd46c4
 
 
 
 
 
4ae50d3
cdd46c4
 
 
 
 
 
4ae50d3
cdd46c4
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4ae50d3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
6a8b0a4
4ae50d3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
8a85c0d
fe32929
7a72693
523b94a
8a85c0d
fe32929
 
 
 
 
af5342c
fe32929
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/quicksrnetlarge/web-assets/model_demo.png)

# QuickSRNetLarge: Optimized for Qualcomm Devices

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

This is based on the implementation of QuickSRNetLarge 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/quicksrnetlarge) 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/quicksrnetlarge/releases/v0.59.0/quicksrnetlarge-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/quicksrnetlarge/releases/v0.59.0/quicksrnetlarge-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/quicksrnetlarge/releases/v0.59.0/quicksrnetlarge-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/quicksrnetlarge/releases/v0.59.0/quicksrnetlarge-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/quicksrnetlarge/releases/v0.59.0/quicksrnetlarge-tflite-float.zip)
| TFLITE | w8a8 | Universal | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/quicksrnetlarge/releases/v0.59.0/quicksrnetlarge-tflite-w8a8.zip)

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


### 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/quicksrnetlarge) 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 [QuickSRNetLarge on GitHub](https://github.com/qualcomm/ai-hub-models/blob/v0.59.0/src/qai_hub_models/models/quicksrnetlarge) for usage instructions.

## Model Details

**Model Type:** Model_use_case.super_resolution

**Model Stats:**
- Model checkpoint: quicksrnet_large_3x_checkpoint
- Input resolution: 128x128
- Number of parameters: 436K
- Model size (float): 1.67 MB
- Model size (w8a8): 462 KB

## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| QuickSRNetLarge | ONNX | float | Snapdragon® X2 Elite | 1.195 ms | 9 - 9 MB | NPU
| QuickSRNetLarge | ONNX | float | Snapdragon® X Elite | 1.964 ms | 8 - 8 MB | NPU
| QuickSRNetLarge | ONNX | float | Snapdragon® 8 Gen 3 Mobile | 1.431 ms | 0 - 38 MB | NPU
| QuickSRNetLarge | ONNX | float | Snapdragon® 8 Gen 1 Mobile | 3.419 ms | 0 - 45 MB | NPU
| QuickSRNetLarge | ONNX | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.908 ms | 0 - 2 MB | NPU
| QuickSRNetLarge | ONNX | float | Qualcomm® QCS8450 | 3.419 ms | 0 - 45 MB | NPU
| QuickSRNetLarge | ONNX | float | Qualcomm® Dragonwing™ IQ-9075 | 3.33 ms | 8 - 10 MB | NPU
| QuickSRNetLarge | ONNX | float | Qualcomm® Dragonwing™ IQ-X7181 | 1.964 ms | 8 - 8 MB | NPU
| QuickSRNetLarge | ONNX | float | Qualcomm® Dragonwing™ Q-8750 | 1.202 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | ONNX | float | Snapdragon® 8 Elite Mobile | 1.202 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.943 ms | 0 - 23 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Snapdragon® X2 Elite | 0.355 ms | 3 - 3 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Snapdragon® X Elite | 0.791 ms | 3 - 3 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.509 ms | 37 - 74 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Snapdragon® 8 Gen 1 Mobile | 1.073 ms | 1 - 41 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Qualcomm® Dragonwing™ QCS6490 | 3.141 ms | 0 - 3 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 0.78 ms | 0 - 2 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Qualcomm® QCS8450 | 1.073 ms | 1 - 41 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Qualcomm® Dragonwing™ IQ-9075 | 1.058 ms | 0 - 3 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Qualcomm® Dragonwing™ IQ-X7181 | 0.791 ms | 3 - 3 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Qualcomm® Dragonwing™ Q-8750 | 0.383 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Snapdragon® 8 Elite Mobile | 0.383 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | ONNX | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.335 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Snapdragon® X2 Elite | 1.237 ms | 0 - 0 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Snapdragon® X Elite | 2.008 ms | 0 - 0 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Snapdragon® 8 Gen 3 Mobile | 1.283 ms | 0 - 36 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Snapdragon® 8 Gen 1 Mobile | 3.276 ms | 0 - 43 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Qualcomm® Dragonwing™ QCS8275 | 11.834 ms | 0 - 23 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 1.807 ms | 0 - 2 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Qualcomm® SA8775P | 3.406 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Qualcomm® SA8650P | 3.406 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Qualcomm® SA8255P | 3.406 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Qualcomm® QCS8450 | 3.276 ms | 0 - 43 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-9075 | 3.243 ms | 0 - 5 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Qualcomm® Dragonwing™ IQ-X7181 | 2.008 ms | 0 - 0 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Qualcomm® Dragonwing™ Q-8750 | 1.041 ms | 0 - 29 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Qualcomm® SA7255P | 11.834 ms | 0 - 23 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Qualcomm® SA8295P | 3.937 ms | 0 - 20 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Snapdragon® 8 Elite Mobile | 1.041 ms | 0 - 29 MB | NPU
| QuickSRNetLarge | QNN_DLC | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.862 ms | 0 - 25 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Snapdragon® X2 Elite | 0.401 ms | 0 - 0 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Snapdragon® X Elite | 0.761 ms | 0 - 0 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.44 ms | 0 - 34 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Snapdragon® 8 Gen 1 Mobile | 0.817 ms | 0 - 39 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ QCS6490 | 3.075 ms | 2 - 4 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ QCS8275 | 2.079 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 0.664 ms | 0 - 4 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® SA8775P | 0.848 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® SA8650P | 0.848 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® SA8255P | 0.848 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® QCS8450 | 0.817 ms | 0 - 39 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ IQ-9075 | 0.75 ms | 2 - 4 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ IQ-X7181 | 0.761 ms | 0 - 0 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ Q-6690 | 8.321 ms | 0 - 135 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ Q-7790 | 0.913 ms | 0 - 22 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® Dragonwing™ Q-8750 | 0.365 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® SA7255P | 2.079 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Qualcomm® SA8295P | 1.408 ms | 0 - 21 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Snapdragon® 8 Elite Mobile | 0.365 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.273 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | QNN_DLC | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.913 ms | 0 - 22 MB | NPU
| QuickSRNetLarge | TFLITE | float | Snapdragon® 8 Gen 3 Mobile | 1.453 ms | 0 - 39 MB | NPU
| QuickSRNetLarge | TFLITE | float | Snapdragon® 8 Gen 1 Mobile | 3.516 ms | 0 - 44 MB | NPU
| QuickSRNetLarge | TFLITE | float | Qualcomm® Dragonwing™ QCS8275 | 12.216 ms | 3 - 28 MB | NPU
| QuickSRNetLarge | TFLITE | float | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 2.148 ms | 0 - 1 MB | NPU
| QuickSRNetLarge | TFLITE | float | Qualcomm® SA8775P | 3.82 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | TFLITE | float | Qualcomm® SA8650P | 3.82 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | TFLITE | float | Qualcomm® SA8255P | 3.82 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | TFLITE | float | Qualcomm® QCS8450 | 3.516 ms | 0 - 44 MB | NPU
| QuickSRNetLarge | TFLITE | float | Qualcomm® Dragonwing™ IQ-9075 | 3.619 ms | 3 - 9 MB | NPU
| QuickSRNetLarge | TFLITE | float | Qualcomm® Dragonwing™ Q-8750 | 1.077 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | TFLITE | float | Qualcomm® SA7255P | 12.216 ms | 3 - 28 MB | NPU
| QuickSRNetLarge | TFLITE | float | Qualcomm® SA8295P | 4.167 ms | 3 - 25 MB | NPU
| QuickSRNetLarge | TFLITE | float | Snapdragon® 8 Elite Mobile | 1.077 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | TFLITE | float | Snapdragon® 8 Elite Gen 5 Mobile | 0.875 ms | 0 - 27 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Snapdragon® 8 Gen 3 Mobile | 0.487 ms | 0 - 35 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Snapdragon® 8 Gen 1 Mobile | 0.912 ms | 1 - 41 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® Dragonwing™ QCS6490 | 3.171 ms | 0 - 3 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® Dragonwing™ QCS8275 | 2.176 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® Dragonwing™ QCS8550 (Proxy) | 0.783 ms | 0 - 4 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® SA8775P | 1.012 ms | 0 - 27 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® SA8650P | 1.012 ms | 0 - 27 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® SA8255P | 1.012 ms | 0 - 27 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® QCS8450 | 0.912 ms | 1 - 41 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® Dragonwing™ IQ-9075 | 0.971 ms | 0 - 3 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® Dragonwing™ Q-6690 | 8.547 ms | 2 - 138 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® Dragonwing™ Q-7790 | 0.97 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® Dragonwing™ Q-8750 | 0.383 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® SA7255P | 2.176 ms | 0 - 26 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Qualcomm® SA8295P | 1.465 ms | 0 - 23 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Snapdragon® 8 Elite Mobile | 0.383 ms | 0 - 24 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Snapdragon® 8 Elite Gen 5 Mobile | 0.301 ms | 0 - 27 MB | NPU
| QuickSRNetLarge | TFLITE | w8a8 | Snapdragon® 7 Gen 4 Mobile | 0.97 ms | 0 - 24 MB | NPU

## License
* The license for the original implementation of QuickSRNetLarge 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).