Hugging Face's logo

qualcomm
/
HRNetPose

Keypoint Detection
PyTorch
TF Lite
ONNX
android
Model card Files Community
qaihm-bot commited on
Commit
3d508b7
·
verified ·
1 Parent(s): d9b6b1d

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +7 -14
README.md CHANGED
@@ -25,15 +25,15 @@ More details on model performance across various devices, can be found
25
  - **Model Type:** Pose estimation
26
  - **Model Stats:**
27
  - Model checkpoint: hrnet_posenet_FP32_state_dict
28
- - Input resolution: 192x256
29
  - Number of parameters: 28.5M
30
  - Model size: 109 MB
31
 
32
 
33
  | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
34
  | ---|---|---|---|---|---|---|---|
35
- | Samsung Galaxy S23 Ultra (Android 13) | Snapdragon® 8 Gen 2 | TFLite | 2.289 ms | 0 - 3 MB | FP16 | NPU | [HRNetPose.tflite](https://huggingface.co/qualcomm/HRNetPose/blob/main/HRNetPose.tflite)
36
- | Samsung Galaxy S23 Ultra (Android 13) | Snapdragon® 8 Gen 2 | QNN Model Library | 2.297 ms | 0 - 57 MB | FP16 | NPU | [HRNetPose.so](https://huggingface.co/qualcomm/HRNetPose/blob/main/HRNetPose.so)
37
 
38
 
39
  ## Installation
@@ -94,17 +94,10 @@ python -m qai_hub_models.models.hrnet_pose.export
94
  ```
95
  Profile Job summary of HRNetPose
96
  --------------------------------------------------
97
- Device: QCS8550 (Proxy) (12)
98
- Estimated Inference Time: 2.29 ms
99
- Estimated Peak Memory Range: 0.02-3.37 MB
100
- Compute Units: NPU (514) | Total (514)
101
-
102
- Profile Job summary of HRNetPose
103
- --------------------------------------------------
104
- Device: QCS8550 (Proxy) (12)
105
- Estimated Inference Time: 2.29 ms
106
- Estimated Peak Memory Range: 0.58-56.72 MB
107
- Compute Units: NPU (745) | Total (745)
108
 
109
 
110
  ```
 
25
  - **Model Type:** Pose estimation
26
  - **Model Stats:**
27
  - Model checkpoint: hrnet_posenet_FP32_state_dict
28
+ - Input resolution: 256x192
29
  - Number of parameters: 28.5M
30
  - Model size: 109 MB
31
 
32
 
33
  | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
34
  | ---|---|---|---|---|---|---|---|
35
+ | Samsung Galaxy S23 Ultra (Android 13) | Snapdragon® 8 Gen 2 | TFLite | 2.818 ms | 0 - 3 MB | FP16 | NPU | [HRNetPose.tflite](https://huggingface.co/qualcomm/HRNetPose/blob/main/HRNetPose.tflite)
36
+ | Samsung Galaxy S23 Ultra (Android 13) | Snapdragon® 8 Gen 2 | QNN Model Library | 2.886 ms | 0 - 20 MB | FP16 | NPU | [HRNetPose.so](https://huggingface.co/qualcomm/HRNetPose/blob/main/HRNetPose.so)
37
 
38
 
39
  ## Installation
 
94
  ```
95
  Profile Job summary of HRNetPose
96
  --------------------------------------------------
97
+ Device: Snapdragon X Elite CRD (11)
98
+ Estimated Inference Time: 3.16 ms
99
+ Estimated Peak Memory Range: 0.56-0.56 MB
100
+ Compute Units: NPU (747) | Total (747)
 
 
 
 
 
 
 
101
 
102
 
103
  ```