Hugging Face's logo

qualcomm
/
QuickSRNetMedium

Image-to-Image
PyTorch
LiteRT
android
Model card Files
xet
Community
QuickSRNetMedium
File size: 20,704 Bytes 
d73fff9
 
 
 
 
cbc1c2e
d73fff9
 
 
5d2cbcb
d73fff9
 
 
 
b997c08
d73fff9
 
6492cc4
b997c08
 
d73fff9
 
 
 
 
4a22578
d73fff9
 
14c0977
d73fff9
a690e30
71a2e90
33efca8
 
 
a690e30
14c0977
d3d01c5
9f26191
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0a472cb
a690e30
d73fff9
 
 
 
 
 
56de2b6
d73fff9
 
 
 
 
2584ed3
d73fff9
2584ed3
d73fff9
 
 
 
 
 
 
2584ed3
d73fff9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4a22578
a690e30
 
d73fff9
 
a690e30
d73fff9
 
 
 
 
 
 
 
 
 
 
 
7ce5592
d73fff9
 
7ce5592
d73fff9
 
d5b8ad9
d73fff9
7ce5592
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d73fff9
 
 
 
 
 
 
 
 
 
 
 
13c4737
 
 
005b34e
d73fff9
 
 
 
 
 
 
 
 
13c4737
 
 
 
005b34e
d73fff9
 
 
 
 
 
2584ed3
d73fff9
 
a690e30
d73fff9
 
 
 
 
761b35e
d73fff9
 
 
 
 
761b35e
d73fff9
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d3d01c5
d73fff9
56de2b6
 
d3d01c5
 
 
d73fff9
 
 
 
 
d3d01c5
 
d73fff9
e241025
d73fff9
 
 
 
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
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
 
---
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/quicksrnetmedium/web-assets/model_demo.png)

# QuickSRNetMedium: Optimized for Mobile Deployment
## Upscale images and remove image noise


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

This model is an implementation of QuickSRNetMedium found [here](https://github.com/quic/aimet-model-zoo/tree/develop/aimet_zoo_torch/quicksrnet).


This repository provides scripts to run QuickSRNetMedium on Qualcomm® devices.
More details on model performance across various devices, can be found
[here](https://aihub.qualcomm.com/models/quicksrnetmedium).



### Model Details

- **Model Type:** Model_use_case.super_resolution
- **Model Stats:**
  - Model checkpoint: quicksrnet_medium_3x_checkpoint
  - Input resolution: 128x128
  - Number of parameters: 61.0K
  - Model size (float): 243 KB
  - Model size (w8a8): 73.9 KB

| Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
|---|---|---|---|---|---|---|---|---|
| QuickSRNetMedium | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 3.001 ms | 1 - 18 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 2.476 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 1.285 ms | 0 - 22 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 1.148 ms | 0 - 28 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 1.058 ms | 0 - 5 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 0.78 ms | 0 - 7 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | ONNX | 1.195 ms | 0 - 4 MB | NPU | [QuickSRNetMedium.onnx.zip](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.onnx.zip) |
| QuickSRNetMedium | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 1.543 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 4.263 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 3.001 ms | 1 - 18 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 2.476 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 1.069 ms | 0 - 6 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 0.779 ms | 0 - 6 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 1.757 ms | 0 - 21 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 1.562 ms | 0 - 22 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 1.07 ms | 0 - 5 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 0.78 ms | 0 - 6 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 1.543 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 4.263 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 0.666 ms | 0 - 29 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 0.494 ms | 0 - 31 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 0.776 ms | 0 - 22 MB | NPU | [QuickSRNetMedium.onnx.zip](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.onnx.zip) |
| QuickSRNetMedium | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 0.486 ms | 0 - 24 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.381 ms | 0 - 21 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | ONNX | 0.616 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.onnx.zip](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.onnx.zip) |
| QuickSRNetMedium | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 0.43 ms | 0 - 18 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.tflite) |
| QuickSRNetMedium | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.369 ms | 0 - 21 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | ONNX | 0.557 ms | 3 - 20 MB | NPU | [QuickSRNetMedium.onnx.zip](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.onnx.zip) |
| QuickSRNetMedium | float | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 0.892 ms | 1 - 1 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.dlc) |
| QuickSRNetMedium | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 1.113 ms | 8 - 8 MB | NPU | [QuickSRNetMedium.onnx.zip](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium.onnx.zip) |
| QuickSRNetMedium | w8a8 | Dragonwing RB3 Gen 2 Vision Kit | Qualcomm® QCS6490 | TFLITE | 1.168 ms | 0 - 3 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | Dragonwing RB3 Gen 2 Vision Kit | Qualcomm® QCS6490 | QNN_DLC | 0.957 ms | 0 - 97 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 0.953 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | QNN_DLC | 0.879 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 0.51 ms | 1 - 24 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | QNN_DLC | 0.518 ms | 0 - 23 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 0.435 ms | 0 - 10 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | QNN_DLC | 0.363 ms | 0 - 10 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 0.651 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | QNN_DLC | 0.543 ms | 0 - 16 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | RB5 (Proxy) | Qualcomm® QCS8250 (Proxy) | TFLITE | 13.392 ms | 1 - 3 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 0.953 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | SA7255P ADP | Qualcomm® SA7255P | QNN_DLC | 0.879 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 0.437 ms | 0 - 9 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | QNN_DLC | 0.34 ms | 0 - 10 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 0.853 ms | 0 - 21 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | SA8295P ADP | Qualcomm® SA8295P | QNN_DLC | 0.733 ms | 0 - 21 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 0.438 ms | 0 - 9 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | QNN_DLC | 0.363 ms | 0 - 10 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 0.651 ms | 0 - 17 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | SA8775P ADP | Qualcomm® SA8775P | QNN_DLC | 0.543 ms | 0 - 16 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 0.274 ms | 0 - 32 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | QNN_DLC | 0.22 ms | 0 - 29 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | TFLITE | 0.233 ms | 0 - 20 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | Samsung Galaxy S25 | Snapdragon® 8 Elite For Galaxy Mobile | QNN_DLC | 0.175 ms | 0 - 20 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | TFLITE | 0.463 ms | 0 - 22 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | Snapdragon 7 Gen 4 QRD | Snapdragon® 7 Gen 4 Mobile | QNN_DLC | 0.376 ms | 0 - 20 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | TFLITE | 0.175 ms | 0 - 19 MB | NPU | [QuickSRNetMedium.tflite](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.tflite) |
| QuickSRNetMedium | w8a8 | Snapdragon 8 Elite Gen 5 QRD | Snapdragon® 8 Elite Gen5 Mobile | QNN_DLC | 0.148 ms | 0 - 19 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |
| QuickSRNetMedium | w8a8 | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN_DLC | 0.447 ms | 6 - 6 MB | NPU | [QuickSRNetMedium.dlc](https://huggingface.co/qualcomm/QuickSRNetMedium/blob/main/QuickSRNetMedium_w8a8.dlc) |




## Installation


Install the package via pip:
```bash
pip install qai-hub-models
```


## Configure Qualcomm® AI Hub Workbench to run this model on a cloud-hosted device

Sign-in to [Qualcomm® AI Hub Workbench](https://workbench.aihub.qualcomm.com/) with your
Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.

With this API token, you can configure your client to run models on the cloud
hosted devices.
```bash
qai-hub configure --api_token API_TOKEN
```
Navigate to [docs](https://workbench.aihub.qualcomm.com/docs/) for more information.



## Demo off target

The package contains a simple end-to-end demo that downloads pre-trained
weights and runs this model on a sample input.

```bash
python -m qai_hub_models.models.quicksrnetmedium.demo
```

The above demo runs a reference implementation of pre-processing, model
inference, and post processing.

**NOTE**: If you want running in a Jupyter Notebook or Google Colab like
environment, please add the following to your cell (instead of the above).
```
%run -m qai_hub_models.models.quicksrnetmedium.demo
```


### Run model on a cloud-hosted device

In addition to the demo, you can also run the model on a cloud-hosted Qualcomm®
device. This script does the following:
* Performance check on-device on a cloud-hosted device
* Downloads compiled assets that can be deployed on-device for Android.
* Accuracy check between PyTorch and on-device outputs.

```bash
python -m qai_hub_models.models.quicksrnetmedium.export
```



## How does this work?

This [export script](https://aihub.qualcomm.com/models/quicksrnetmedium/qai_hub_models/models/QuickSRNetMedium/export.py)
leverages [Qualcomm® AI Hub](https://aihub.qualcomm.com/) to optimize, validate, and deploy this model
on-device. Lets go through each step below in detail:

Step 1: **Compile model for on-device deployment**

To compile a PyTorch model for on-device deployment, we first trace the model
in memory using the `jit.trace` and then call the `submit_compile_job` API.

```python
import torch

import qai_hub as hub
from qai_hub_models.models.quicksrnetmedium import Model

# Load the model
torch_model = Model.from_pretrained()

# Device
device = hub.Device("Samsung Galaxy S25")

# Trace model
input_shape = torch_model.get_input_spec()
sample_inputs = torch_model.sample_inputs()

pt_model = torch.jit.trace(torch_model, [torch.tensor(data[0]) for _, data in sample_inputs.items()])

# Compile model on a specific device
compile_job = hub.submit_compile_job(
    model=pt_model,
    device=device,
    input_specs=torch_model.get_input_spec(),
)

# Get target model to run on-device
target_model = compile_job.get_target_model()

```


Step 2: **Performance profiling on cloud-hosted device**

After compiling models from step 1. Models can be profiled model on-device using the
`target_model`. Note that this scripts runs the model on a device automatically
provisioned in the cloud.  Once the job is submitted, you can navigate to a
provided job URL to view a variety of on-device performance metrics.
```python
profile_job = hub.submit_profile_job(
    model=target_model,
    device=device,
)
        
```

Step 3: **Verify on-device accuracy**

To verify the accuracy of the model on-device, you can run on-device inference
on sample input data on the same cloud hosted device.
```python
input_data = torch_model.sample_inputs()
inference_job = hub.submit_inference_job(
    model=target_model,
    device=device,
    inputs=input_data,
)
    on_device_output = inference_job.download_output_data()

```
With the output of the model, you can compute like PSNR, relative errors or
spot check the output with expected output.

**Note**: This on-device profiling and inference requires access to Qualcomm®
AI Hub Workbench. [Sign up for access](https://myaccount.qualcomm.com/signup).



## Run demo on a cloud-hosted device

You can also run the demo on-device.

```bash
python -m qai_hub_models.models.quicksrnetmedium.demo --eval-mode on-device
```

**NOTE**: If you want running in a Jupyter Notebook or Google Colab like
environment, please add the following to your cell (instead of the above).
```
%run -m qai_hub_models.models.quicksrnetmedium.demo -- --eval-mode on-device
```


## Deploying compiled model to Android


The models can be deployed using multiple runtimes:
- TensorFlow Lite (`.tflite` export): [This
  tutorial](https://www.tensorflow.org/lite/android/quickstart) provides a
  guide to deploy the .tflite model in an Android application.


- QNN (`.so` export ): This [sample
  app](https://docs.qualcomm.com/bundle/publicresource/topics/80-63442-50/sample_app.html)
provides instructions on how to use the `.so` shared library  in an Android application.


## View on Qualcomm® AI Hub
Get more details on QuickSRNetMedium's performance across various devices [here](https://aihub.qualcomm.com/models/quicksrnetmedium).
Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)


## License
* The license for the original implementation of QuickSRNetMedium can be found
  [here](https://github.com/quic/aimet-model-zoo/blob/develop/LICENSE.pdf).
* The license for the compiled assets for on-device deployment can be found [here](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/Qualcomm+AI+Hub+Proprietary+License.pdf)



## 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:[email protected]).