FFNet-122NS-LowRes: Optimized for Qualcomm Devices

FFNet-122NS-LowRes is a "fuss-free network" that segments street scene images with per-pixel classes like road, sidewalk, and pedestrian. Trained on the Cityscapes dataset.

This is based on the implementation of FFNet-122NS-LowRes 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 2.42, ONNX Runtime 1.24.1	Download
ONNX	w8a8	Universal	QAIRT 2.42, ONNX Runtime 1.24.1	Download
QNN_DLC	float	Universal	QAIRT 2.43	Download
QNN_DLC	w8a8	Universal	QAIRT 2.43	Download
TFLITE	float	Universal	QAIRT 2.43, TFLite 2.17.0	Download
TFLITE	w8a8	Universal	QAIRT 2.43, TFLite 2.17.0	Download

For more device-specific assets and performance metrics, visit FFNet-122NS-LowRes 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 FFNet-122NS-LowRes on GitHub for usage instructions.

Model Details

Model Type: Model_use_case.semantic_segmentation

Model Stats:

• Model checkpoint: ffnet122NS_CCC_cityscapes_state_dict_quarts_pre_down
• Input resolution: 1024x512
• Number of output classes: 19
• Number of parameters: 32.1M
• Model size (float): 123 MB
• Model size (w8a8): 31.3 MB

Performance Summary

Model	Runtime	Precision	Chipset	Inference Time (ms)	Peak Memory Range (MB)	Primary Compute Unit
FFNet-122NS-LowRes	ONNX	float	Snapdragon® 8 Elite Gen 5 Mobile	3.507 ms	2 - 181 MB	NPU
FFNet-122NS-LowRes	ONNX	float	Snapdragon® X2 Elite	3.603 ms	57 - 57 MB	NPU
FFNet-122NS-LowRes	ONNX	float	Snapdragon® X Elite	9.041 ms	56 - 56 MB	NPU
FFNet-122NS-LowRes	ONNX	float	Snapdragon® 8 Gen 3 Mobile	7.967 ms	7 - 228 MB	NPU
FFNet-122NS-LowRes	ONNX	float	Qualcomm® QCS8550 (Proxy)	9.893 ms	0 - 60 MB	NPU
FFNet-122NS-LowRes	ONNX	float	Qualcomm® QCS9075	10.232 ms	6 - 14 MB	NPU
FFNet-122NS-LowRes	ONNX	float	Snapdragon® 8 Elite For Galaxy Mobile	6.339 ms	3 - 181 MB	NPU
FFNet-122NS-LowRes	ONNX	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	1.321 ms	0 - 87 MB	NPU
FFNet-122NS-LowRes	ONNX	w8a8	Snapdragon® X2 Elite	1.281 ms	30 - 30 MB	NPU
FFNet-122NS-LowRes	ONNX	w8a8	Snapdragon® X Elite	2.818 ms	30 - 30 MB	NPU
FFNet-122NS-LowRes	ONNX	w8a8	Snapdragon® 8 Gen 3 Mobile	1.884 ms	0 - 156 MB	NPU
FFNet-122NS-LowRes	ONNX	w8a8	Qualcomm® QCS6490	91.995 ms	54 - 149 MB	CPU
FFNet-122NS-LowRes	ONNX	w8a8	Qualcomm® QCS8550 (Proxy)	2.614 ms	0 - 33 MB	NPU
FFNet-122NS-LowRes	ONNX	w8a8	Qualcomm® QCS9075	3.127 ms	1 - 4 MB	NPU
FFNet-122NS-LowRes	ONNX	w8a8	Qualcomm® QCM6690	85.686 ms	51 - 64 MB	CPU
FFNet-122NS-LowRes	ONNX	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	1.469 ms	0 - 93 MB	NPU
FFNet-122NS-LowRes	ONNX	w8a8	Snapdragon® 7 Gen 4 Mobile	85.442 ms	59 - 73 MB	CPU
FFNet-122NS-LowRes	QNN_DLC	float	Snapdragon® 8 Elite Gen 5 Mobile	5.646 ms	6 - 173 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	float	Snapdragon® X2 Elite	6.338 ms	6 - 6 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	float	Snapdragon® X Elite	12.985 ms	6 - 6 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	float	Snapdragon® 8 Gen 3 Mobile	8.777 ms	6 - 208 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	float	Qualcomm® QCS8275 (Proxy)	39.601 ms	2 - 162 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	float	Qualcomm® QCS8550 (Proxy)	12.817 ms	6 - 8 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	float	Qualcomm® SA8775P	16.048 ms	1 - 163 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	float	Qualcomm® QCS9075	16.895 ms	6 - 14 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	float	Qualcomm® QCS8450 (Proxy)	29.95 ms	6 - 193 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	float	Qualcomm® SA7255P	39.601 ms	2 - 162 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	float	Qualcomm® SA8295P	17.722 ms	1 - 159 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	float	Snapdragon® 8 Elite For Galaxy Mobile	6.971 ms	0 - 163 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	1.797 ms	2 - 80 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Snapdragon® X2 Elite	1.966 ms	2 - 2 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Snapdragon® X Elite	4.833 ms	2 - 2 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Snapdragon® 8 Gen 3 Mobile	3.177 ms	0 - 128 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Qualcomm® QCS6490	13.915 ms	1 - 5 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Qualcomm® QCS8275 (Proxy)	9.261 ms	1 - 76 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Qualcomm® QCS8550 (Proxy)	4.5 ms	2 - 3 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Qualcomm® SA8775P	4.984 ms	2 - 79 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Qualcomm® QCS9075	5.874 ms	1 - 5 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Qualcomm® QCM6690	25.738 ms	2 - 199 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Qualcomm® QCS8450 (Proxy)	6.878 ms	2 - 126 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Qualcomm® SA7255P	9.261 ms	1 - 76 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Qualcomm® SA8295P	6.045 ms	1 - 75 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	2.178 ms	2 - 75 MB	NPU
FFNet-122NS-LowRes	QNN_DLC	w8a8	Snapdragon® 7 Gen 4 Mobile	5.733 ms	2 - 195 MB	NPU
FFNet-122NS-LowRes	TFLITE	float	Snapdragon® 8 Elite Gen 5 Mobile	5.404 ms	1 - 195 MB	NPU
FFNet-122NS-LowRes	TFLITE	float	Snapdragon® 8 Gen 3 Mobile	8.618 ms	0 - 275 MB	NPU
FFNet-122NS-LowRes	TFLITE	float	Qualcomm® QCS8275 (Proxy)	39.92 ms	1 - 192 MB	NPU
FFNet-122NS-LowRes	TFLITE	float	Qualcomm® QCS8550 (Proxy)	12.734 ms	1 - 3 MB	NPU
FFNet-122NS-LowRes	TFLITE	float	Qualcomm® SA8775P	16.0 ms	1 - 192 MB	NPU
FFNet-122NS-LowRes	TFLITE	float	Qualcomm® QCS9075	16.933 ms	0 - 70 MB	NPU
FFNet-122NS-LowRes	TFLITE	float	Qualcomm® QCS8450 (Proxy)	30.28 ms	1 - 271 MB	NPU
FFNet-122NS-LowRes	TFLITE	float	Qualcomm® SA7255P	39.92 ms	1 - 192 MB	NPU
FFNet-122NS-LowRes	TFLITE	float	Qualcomm® SA8295P	17.976 ms	1 - 189 MB	NPU
FFNet-122NS-LowRes	TFLITE	float	Snapdragon® 8 Elite For Galaxy Mobile	6.857 ms	0 - 191 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Snapdragon® 8 Elite Gen 5 Mobile	1.314 ms	0 - 75 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Snapdragon® 8 Gen 3 Mobile	1.862 ms	0 - 131 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Qualcomm® QCS6490	8.999 ms	0 - 35 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Qualcomm® QCS8275 (Proxy)	5.945 ms	0 - 71 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Qualcomm® QCS8550 (Proxy)	2.602 ms	0 - 2 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Qualcomm® SA8775P	3.087 ms	0 - 74 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Qualcomm® QCS9075	3.056 ms	0 - 35 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Qualcomm® QCM6690	23.291 ms	0 - 197 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Qualcomm® QCS8450 (Proxy)	3.259 ms	0 - 129 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Qualcomm® SA7255P	5.945 ms	0 - 71 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Qualcomm® SA8295P	3.874 ms	0 - 69 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Snapdragon® 8 Elite For Galaxy Mobile	1.441 ms	0 - 76 MB	NPU
FFNet-122NS-LowRes	TFLITE	w8a8	Snapdragon® 7 Gen 4 Mobile	3.695 ms	0 - 188 MB	NPU

License

• The license for the original implementation of FFNet-122NS-LowRes can be found here.

References

• Simple and Efficient Architectures for Semantic Segmentation
• Source Model Implementation

Community

• Join our AI Hub Slack community to collaborate, post questions and learn more about on-device AI.
• For questions or feedback please reach out to us.