Hugging Face's logo

qualcomm
/
LiteHRNet

Keypoint Detection
PyTorch
LiteRT
ONNX
android
Model card Files Community
qaihm-bot commited on
Commit
ee506f0
·
verified ·
1 Parent(s): 2670ee1

v0.32.0

Browse files

See https://github.com/quic/ai-hub-models/releases/v0.32.0 for changelog.

Files changed (6) hide show
  1. .gitattributes +1 -0
  2. DEPLOYMENT_MODEL_LICENSE.pdf +3 -0
  3. LICENSE +2 -0
  4. LiteHRNet.onnx +2 -2
  5. LiteHRNet.tflite +2 -2
  6. README.md +20 -20
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ DEPLOYMENT_MODEL_LICENSE.pdf filter=lfs diff=lfs merge=lfs -text
DEPLOYMENT_MODEL_LICENSE.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4409f93b0e82531303b3e10f52f1fdfb56467a25f05b7441c6bbd8bb8a64b42c
3
+ size 109629
LICENSE ADDED
@@ -0,0 +1,2 @@
 
 
 
1
+ The license of the original trained model can be found at https://github.com/HRNet/Lite-HRNet/blob/hrnet/LICENSE.
2
+ The license for the deployable model files (.tflite, .onnx, .dlc, .bin, etc.) can be found in DEPLOYMENT_MODEL_LICENSE.pdf.
LiteHRNet.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:7bcae4471ad42875a330144859e316cbce54533f9b5f927509336861de726faf
3
- size 4863249
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0763707b735a472207c19ebdb98a8fa15edd4a2ea38658e84f64d1d17b514678
3
+ size 4897708
LiteHRNet.tflite CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:1c8618fa2ceb767684ebd0dc7aba3e1dc8e1594dbdeccd848c7a635e2c25f539
3
- size 4706720
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2df61b0d03f0b42bab8e0a4e4b383458dd1a0d918ecf3e99d7b5f10203ea3abc
3
+ size 4706560
README.md CHANGED
@@ -29,26 +29,26 @@ More details on model performance across various devices, can be found
29
  - **Model Stats:**
30
  - Input resolution: 256x192
31
  - Number of parameters: 1.11M
32
- - Model size: 4.56 MB
33
 
34
  | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
35
  |---|---|---|---|---|---|---|---|---|
36
- | LiteHRNet | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 8.874 ms | 0 - 65 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
37
- | LiteHRNet | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 5.25 ms | 0 - 67 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
38
- | LiteHRNet | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 4.584 ms | 0 - 23 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
39
- | LiteHRNet | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 5.625 ms | 0 - 65 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
40
- | LiteHRNet | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 8.874 ms | 0 - 65 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
41
- | LiteHRNet | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 4.608 ms | 0 - 22 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
42
- | LiteHRNet | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 6.566 ms | 0 - 62 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
43
- | LiteHRNet | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 4.631 ms | 0 - 15 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
44
- | LiteHRNet | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 5.625 ms | 0 - 65 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
45
- | LiteHRNet | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 4.654 ms | 0 - 22 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
46
- | LiteHRNet | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 6.099 ms | 0 - 23 MB | NPU | [LiteHRNet.onnx](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.onnx) |
47
- | LiteHRNet | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 2.787 ms | 0 - 73 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
48
- | LiteHRNet | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 3.754 ms | 0 - 61 MB | NPU | [LiteHRNet.onnx](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.onnx) |
49
- | LiteHRNet | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 2.867 ms | 0 - 64 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
50
- | LiteHRNet | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 4.066 ms | 1 - 54 MB | NPU | [LiteHRNet.onnx](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.onnx) |
51
- | LiteHRNet | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 6.464 ms | 4 - 4 MB | NPU | [LiteHRNet.onnx](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.onnx) |
52
 
53
 
54
 
@@ -112,10 +112,10 @@ Profiling Results
112
  LiteHRNet
113
  Device : cs_8275 (ANDROID 14)
114
  Runtime : TFLITE
115
- Estimated inference time (ms) : 8.9
116
  Estimated peak memory usage (MB): [0, 65]
117
- Total # Ops : 1113
118
- Compute Unit(s) : npu (1111 ops) gpu (0 ops) cpu (2 ops)
119
  ```
120
 
121
 
 
29
  - **Model Stats:**
30
  - Input resolution: 256x192
31
  - Number of parameters: 1.11M
32
+ - Model size (float): 4.49 MB
33
 
34
  | Model | Precision | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit | Target Model
35
  |---|---|---|---|---|---|---|---|---|
36
+ | LiteHRNet | float | QCS8275 (Proxy) | Qualcomm® QCS8275 (Proxy) | TFLITE | 8.812 ms | 0 - 65 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
37
+ | LiteHRNet | float | QCS8450 (Proxy) | Qualcomm® QCS8450 (Proxy) | TFLITE | 5.228 ms | 0 - 68 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
38
+ | LiteHRNet | float | QCS8550 (Proxy) | Qualcomm® QCS8550 (Proxy) | TFLITE | 4.51 ms | 0 - 26 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
39
+ | LiteHRNet | float | QCS9075 (Proxy) | Qualcomm® QCS9075 (Proxy) | TFLITE | 5.542 ms | 0 - 64 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
40
+ | LiteHRNet | float | SA7255P ADP | Qualcomm® SA7255P | TFLITE | 8.812 ms | 0 - 65 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
41
+ | LiteHRNet | float | SA8255 (Proxy) | Qualcomm® SA8255P (Proxy) | TFLITE | 4.553 ms | 0 - 21 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
42
+ | LiteHRNet | float | SA8295P ADP | Qualcomm® SA8295P | TFLITE | 6.483 ms | 0 - 62 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
43
+ | LiteHRNet | float | SA8650 (Proxy) | Qualcomm® SA8650P (Proxy) | TFLITE | 4.543 ms | 0 - 23 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
44
+ | LiteHRNet | float | SA8775P ADP | Qualcomm® SA8775P | TFLITE | 5.542 ms | 0 - 64 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
45
+ | LiteHRNet | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | TFLITE | 4.587 ms | 0 - 23 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
46
+ | LiteHRNet | float | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 Mobile | ONNX | 6.02 ms | 0 - 21 MB | NPU | [LiteHRNet.onnx](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.onnx) |
47
+ | LiteHRNet | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | TFLITE | 2.757 ms | 0 - 71 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
48
+ | LiteHRNet | float | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 Mobile | ONNX | 3.707 ms | 0 - 59 MB | NPU | [LiteHRNet.onnx](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.onnx) |
49
+ | LiteHRNet | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | TFLITE | 2.816 ms | 0 - 64 MB | NPU | [LiteHRNet.tflite](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.tflite) |
50
+ | LiteHRNet | float | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite Mobile | ONNX | 4.03 ms | 1 - 52 MB | NPU | [LiteHRNet.onnx](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.onnx) |
51
+ | LiteHRNet | float | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 6.494 ms | 4 - 4 MB | NPU | [LiteHRNet.onnx](https://huggingface.co/qualcomm/LiteHRNet/blob/main/LiteHRNet.onnx) |
52
 
53
 
54
 
 
112
  LiteHRNet
113
  Device : cs_8275 (ANDROID 14)
114
  Runtime : TFLITE
115
+ Estimated inference time (ms) : 8.8
116
  Estimated peak memory usage (MB): [0, 65]
117
+ Total # Ops : 1112
118
+ Compute Unit(s) : npu (1110 ops) gpu (0 ops) cpu (2 ops)
119
  ```
120
 
121