---
library_name: pytorch
license: other
tags:
- android
pipeline_tag: other

---

![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/web-assets/model_demo.png)

# BEVFusion: Optimized for Qualcomm Devices

BeVFusion is a machine learning model for generating a birds eye view represenation from the sensors(cameras) mounted on a vehicle.

This is based on the implementation of BEVFusion found [here](https://github.com/mit-han-lab/bevfusion).
This repository contains pre-exported model files optimized for Qualcomm® devices. You can use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/bevfusion_det) library to export with custom configurations. More details on model performance across various devices, can be found [here](#performance-summary).

Qualcomm AI Hub Models uses [Qualcomm AI Hub Workbench](https://workbench.aihub.qualcomm.com) to compile, profile, and evaluate this model. [Sign up](https://myaccount.qualcomm.com/signup) 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 |
|---|---|---|---|---|
| PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-precompiled_qnn_onnx-float-qualcomm_snapdragon_8_elite_gen5.zip)
| PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-precompiled_qnn_onnx-float-qualcomm_snapdragon_x2_elite.zip)
| PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-precompiled_qnn_onnx-float-qualcomm_snapdragon_x_elite.zip)
| PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-precompiled_qnn_onnx-float-qualcomm_snapdragon_8gen3.zip)
| PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-precompiled_qnn_onnx-float-qualcomm_qcs8550_proxy.zip)
| PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-precompiled_qnn_onnx-float-qualcomm_snapdragon_8_elite_for_galaxy.zip)
| PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | QAIRT 2.42, ONNX Runtime 1.24.3 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-precompiled_qnn_onnx-float-qualcomm_qcs9075.zip)
| QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-qnn_context_binary-float-qualcomm_snapdragon_8_elite_gen5.zip)
| QNN_CONTEXT_BINARY | float | Snapdragon® X2 Elite | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-qnn_context_binary-float-qualcomm_snapdragon_x2_elite.zip)
| QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-qnn_context_binary-float-qualcomm_snapdragon_x_elite.zip)
| QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-qnn_context_binary-float-qualcomm_snapdragon_8gen3.zip)
| QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-qnn_context_binary-float-qualcomm_qcs8550_proxy.zip)
| QNN_CONTEXT_BINARY | float | Qualcomm® SA8775P | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-qnn_context_binary-float-qualcomm_sa8775p.zip)
| QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-qnn_context_binary-float-qualcomm_snapdragon_8_elite_for_galaxy.zip)
| QNN_CONTEXT_BINARY | float | Qualcomm® SA7255P | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-qnn_context_binary-float-qualcomm_sa7255p.zip)
| QNN_CONTEXT_BINARY | float | Qualcomm® SA8295P | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-qnn_context_binary-float-qualcomm_sa8295p.zip)
| QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | QAIRT 2.45 | [Download](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/bevfusion_det/releases/v0.51.0/bevfusion_det-qnn_context_binary-float-qualcomm_qcs9075.zip)

For more device-specific assets and performance metrics, visit **[BEVFusion on Qualcomm® AI Hub](https://aihub.qualcomm.com/models/bevfusion_det)**.


### Option 2: Export with Custom Configurations

Use the [Qualcomm® AI Hub Models](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/bevfusion_det) 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 [BEVFusion on GitHub](https://github.com/qualcomm/ai-hub-models/blob/main/src/qai_hub_models/models/bevfusion_det) for usage instructions.

## Model Details

**Model Type:** Model_use_case.driver_assistance

**Model Stats:**
- Model checkpoint: camera-only-det.pth
- Input resolution: 1 x 6 x 3 x 256 x 704
- Number of parameters: 44M
- Model size: 171 MB

## Performance Summary
| Model | Runtime | Precision | Chipset | Inference Time (ms) | Peak Memory Range (MB) | Primary Compute Unit
|---|---|---|---|---|---|---
| decoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 5.743 ms | 15 - 25 MB | NPU
| decoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 7.014 ms | 24 - 24 MB | NPU
| decoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 13.073 ms | 23 - 23 MB | NPU
| decoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 9.877 ms | 17 - 24 MB | NPU
| decoder | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | 12.632 ms | 4 - 32 MB | NPU
| decoder | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | 24.372 ms | 16 - 19 MB | NPU
| decoder | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 7.685 ms | 12 - 19 MB | NPU
| decoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 5.708 ms | 5 - 14 MB | NPU
| decoder | QNN_CONTEXT_BINARY | float | Snapdragon® X2 Elite | 7.428 ms | 5 - 5 MB | NPU
| decoder | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 13.161 ms | 5 - 5 MB | NPU
| decoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 9.794 ms | 5 - 12 MB | NPU
| decoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8275 (Proxy) | 92.805 ms | 0 - 10 MB | NPU
| decoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | 12.809 ms | 5 - 7 MB | NPU
| decoder | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | 24.65 ms | 5 - 16 MB | NPU
| decoder | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | 7.768 ms | 5 - 18 MB | NPU
| encoder_1 | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 350.268 ms | 53 - 62 MB | NPU
| encoder_1 | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 380.872 ms | 102 - 102 MB | NPU
| encoder_1 | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 685.608 ms | 101 - 101 MB | NPU
| encoder_1 | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 521.229 ms | 33 - 45 MB | NPU
| encoder_1 | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | 730.457 ms | 0 - 116 MB | NPU
| encoder_1 | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | 705.627 ms | 12 - 27 MB | NPU
| encoder_1 | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 422.927 ms | 36 - 49 MB | NPU
| encoder_1 | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 353.268 ms | 12 - 22 MB | NPU
| encoder_1 | QNN_CONTEXT_BINARY | float | Snapdragon® X2 Elite | 391.075 ms | 12 - 12 MB | NPU
| encoder_1 | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 684.197 ms | 12 - 12 MB | NPU
| encoder_1 | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 522.428 ms | 13 - 20 MB | NPU
| encoder_1 | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8275 (Proxy) | 1074.155 ms | 0 - 9 MB | NPU
| encoder_1 | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | 733.049 ms | 13 - 14 MB | NPU
| encoder_1 | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | 704.381 ms | 12 - 43 MB | NPU
| encoder_1 | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | 436.38 ms | 12 - 25 MB | NPU
| encoder_2 | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 2174.83 ms | 49 - 59 MB | NPU
| encoder_2 | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 2371.931 ms | 1058 - 1058 MB | NPU
| encoder_2 | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 3419.981 ms | 1058 - 1058 MB | NPU
| encoder_2 | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 2698.622 ms | 538 - 544 MB | NPU
| encoder_2 | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | 3639.158 ms | 586 - 587 MB | NPU
| encoder_2 | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | 3602.42 ms | 520 - 522 MB | NPU
| encoder_2 | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 2451.979 ms | 591 - 598 MB | NPU
| encoder_2 | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 2193.918 ms | 17 - 27 MB | NPU
| encoder_2 | QNN_CONTEXT_BINARY | float | Snapdragon® X2 Elite | 2350.242 ms | 17 - 17 MB | NPU
| encoder_2 | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 3370.765 ms | 17 - 17 MB | NPU
| encoder_2 | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 2643.753 ms | 17 - 28 MB | NPU
| encoder_2 | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8275 (Proxy) | 5454.655 ms | 16 - 26 MB | NPU
| encoder_2 | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | 3415.223 ms | 17 - 18 MB | NPU
| encoder_2 | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | 3360.631 ms | 17 - 644 MB | NPU
| encoder_2 | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | 2423.043 ms | 17 - 31 MB | NPU
| encoder_3 | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite Gen 5 Mobile | 6.85 ms | 29 - 39 MB | NPU
| encoder_3 | PRECOMPILED_QNN_ONNX | float | Snapdragon® X2 Elite | 7.788 ms | 14 - 14 MB | NPU
| encoder_3 | PRECOMPILED_QNN_ONNX | float | Snapdragon® X Elite | 11.935 ms | 18 - 18 MB | NPU
| encoder_3 | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Gen 3 Mobile | 9.179 ms | 33 - 39 MB | NPU
| encoder_3 | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS8550 (Proxy) | 12.363 ms | 23 - 27 MB | NPU
| encoder_3 | PRECOMPILED_QNN_ONNX | float | Qualcomm® QCS9075 | 15.139 ms | 18 - 39 MB | NPU
| encoder_3 | PRECOMPILED_QNN_ONNX | float | Snapdragon® 8 Elite For Galaxy Mobile | 7.775 ms | 14 - 21 MB | NPU
| encoder_3 | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite Gen 5 Mobile | 6.784 ms | 18 - 28 MB | NPU
| encoder_3 | QNN_CONTEXT_BINARY | float | Snapdragon® X2 Elite | 8.397 ms | 18 - 18 MB | NPU
| encoder_3 | QNN_CONTEXT_BINARY | float | Snapdragon® X Elite | 12.046 ms | 19 - 19 MB | NPU
| encoder_3 | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Gen 3 Mobile | 9.102 ms | 18 - 26 MB | NPU
| encoder_3 | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8275 (Proxy) | 25.327 ms | 18 - 27 MB | NPU
| encoder_3 | QNN_CONTEXT_BINARY | float | Qualcomm® QCS8550 (Proxy) | 12.6 ms | 19 - 20 MB | NPU
| encoder_3 | QNN_CONTEXT_BINARY | float | Qualcomm® QCS9075 | 15.126 ms | 18 - 43 MB | NPU
| encoder_3 | QNN_CONTEXT_BINARY | float | Snapdragon® 8 Elite For Galaxy Mobile | 7.694 ms | 17 - 27 MB | NPU

## License
* The license for the original implementation of BEVFusion can be found
  [here](https://github.com/w-hc/torch_audioset/blob/master/LICENSE).

## References
* [BEVFusion: Multi-Task Multi-Sensor Fusion with Unified Bird's-Eye View Representation](https://arxiv.org/abs/2205.13542)
* [Source Model Implementation](https://github.com/mit-han-lab/bevfusion)

## 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:ai-hub-support@qti.qualcomm.com).