Hugging Peers

HRNet-W48-OCR

☆ Save
by qualcomm image-segmentation pytorch bsd-3-clause
77 downloads 1 likes on Hugging Face updated

Model card

HRNet-W48-OCR: Optimized for Qualcomm Devices

HRNet-W48-OCR is a machine learning model that can segment images from the Cityscape dataset. It has lightweight and hardware-efficient operations and thus delivers significant speedup on diverse hardware platforms

This is based on the implementation of HRNet-W48-OCR found here. This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the Qualcomm® AI Hub Models library to export with custom configurations. More details on model performance across various devices, can be found here.

Qualcomm AI Hub Models uses Qualcomm AI Hub Workbench to compile, profile, and evaluate this model. Sign up 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 2.50, ONNX Runtime 1.30.0 Download
ONNX w8a16 Universal QAIRT 2.50, ONNX Runtime 1.30.0 Download

For more device-specific assets and performance metrics, visit HRNet-W48-OCR on Qualcomm® AI Hub.

Option 2: Export with Custom Configurations

Use the Qualcomm® AI Hub Models 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 HRNet-W48-OCR on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

  • Input resolution: 2048x1024
  • Model checkpoint: hrnet_ocr_cs_8162_torch11.pth
  • Model size (float): 268 MB
  • Model size (w8a16): 70.3 MB
  • Number of output classes: 19
  • Number of parameters: 70.3M

Performance Summary

Model Runtime Precision Chipset Inference Time (ms) Peak Memory Range (MB) Primary Compute Unit
HRNet-W48-OCR ONNX float Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 468.225 ms 21 - 1138 MB NPU
HRNet-W48-OCR ONNX float Snapdragon® 8 Elite For Galaxy Mobile 684.068 ms 22 - 956 MB NPU
HRNet-W48-OCR ONNX float Snapdragon® X2 Elite 534.683 ms 32 - 32 MB NPU
HRNet-W48-OCR ONNX float Snapdragon® X Elite 848.924 ms 132 - 132 MB NPU
HRNet-W48-OCR ONNX float Snapdragon® 8 Gen 3 Mobile 668.713 ms 36 - 2230 MB NPU
HRNet-W48-OCR ONNX float Snapdragon® 8 Gen 1 Mobile 1391.683 ms 26 - 2135 MB NPU
HRNet-W48-OCR ONNX float Qualcomm® Dragonwing™ IQ-8275 1170.997 ms 24 - 52 MB NPU
HRNet-W48-OCR ONNX float Qualcomm® Dragonwing™ QCS8550 (Proxy) 921.95 ms 0 - 135 MB NPU
HRNet-W48-OCR ONNX float Qualcomm® QCS8450 1391.683 ms 26 - 2135 MB NPU
HRNet-W48-OCR ONNX float Qualcomm® Dragonwing™ IQ-9075 1204.134 ms 24 - 51 MB NPU
HRNet-W48-OCR ONNX float Qualcomm® Dragonwing™ IQ-X7181 848.924 ms 132 - 132 MB NPU
HRNet-W48-OCR ONNX float Qualcomm® Dragonwing™ Q-8750 684.068 ms 22 - 956 MB NPU
HRNet-W48-OCR ONNX w8a16 Snapdragon® 8 Elite Gen 5 For Galaxy Mobile 390.463 ms 26 - 2714 MB NPU
HRNet-W48-OCR ONNX w8a16 Snapdragon® 8 Elite For Galaxy Mobile 501.535 ms 21 - 2382 MB NPU
HRNet-W48-OCR ONNX w8a16 Snapdragon® X2 Elite 418.61 ms 34 - 34 MB NPU
HRNet-W48-OCR ONNX w8a16 Snapdragon® X Elite 708.285 ms 89 - 89 MB NPU
HRNet-W48-OCR ONNX w8a16 Snapdragon® 8 Gen 3 Mobile 593.581 ms 34 - 3416 MB NPU
HRNet-W48-OCR ONNX w8a16 Qualcomm® Dragonwing™ IQ-8275 872.011 ms 20 - 35 MB NPU
HRNet-W48-OCR ONNX w8a16 Qualcomm® Dragonwing™ QCS8550 (Proxy) 778.581 ms 0 - 102 MB NPU
HRNet-W48-OCR ONNX w8a16 Qualcomm® QCS8450 1018.392 ms 7 - 2663 MB NPU
HRNet-W48-OCR ONNX w8a16 Qualcomm® Dragonwing™ IQ-9075 859.223 ms 21 - 35 MB NPU
HRNet-W48-OCR ONNX w8a16 Qualcomm® Dragonwing™ IQ-X7181 708.285 ms 89 - 89 MB NPU
HRNet-W48-OCR ONNX w8a16 Qualcomm® Dragonwing™ Q-8750 501.535 ms 21 - 2382 MB NPU

License

  • The license for the original implementation of HRNet-W48-OCR can be found here.

References

Community