qualcomm
/

3D-Deep-BOX

+---
+library_name: pytorch
+license: apache-2.0
+tags:
+- real_time
+- android
+pipeline_tag: object-detection
+---
+![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/deepbox/web-assets/model_demo.png)
+# 3D-Deep-BOX: Optimized for Mobile Deployment
+## Real-time 3D object detection
+3D Deep Box is a machine learning model that predicts 3D bounding boxes and classes of objects in an image.
+This model is an implementation of 3D-Deep-BOX found [here](https://github.com/skhadem/3D-BoundingBox/).
+This repository provides scripts to run 3D-Deep-BOX on Qualcomm® devices.
+More details on model performance across various devices, can be found
+[here](https://aihub.qualcomm.com/models/deepbox).
+### Model Details
+- **Model Type:** Object detection
+- **Model Stats:**
+  - Model checkpoint: YOLOv3-tiny
+  - Input resolution(YOLO): 224x640
+  - Number of parameters(YOLO): 8.85M
+  - Model size(YOLO): 37.3 MB
+  - Input resolution(VGG): 224x224
+  - Number of parameters(VGG): 144M
+  - Model size(VGG): 175.9 MB
+| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model
+|---|---|---|---|---|---|---|---|---|
+| Yolo | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 22.238 ms | 0 - 59 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 2.992 ms | 2 - 4 MB | FP16 | NPU | [3D-Deep-BOX.so](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.so) |
+| Yolo | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 5.749 ms | 0 - 51 MB | FP16 | NPU | [3D-Deep-BOX.onnx](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.onnx) |
+| Yolo | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 16.668 ms | 0 - 39 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 2.085 ms | 0 - 15 MB | FP16 | NPU | [3D-Deep-BOX.so](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.so) |
+| Yolo | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 4.781 ms | 0 - 31 MB | FP16 | NPU | [3D-Deep-BOX.onnx](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.onnx) |
+| Yolo | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 14.794 ms | 0 - 33 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 2.519 ms | 2 - 20 MB | FP16 | NPU | Use Export Script |
+| Yolo | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 4.697 ms | 2 - 27 MB | FP16 | NPU | [3D-Deep-BOX.onnx](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.onnx) |
+| Yolo | SA7255P ADP | SA7255P | TFLITE | 67.992 ms | 0 - 26 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | SA7255P ADP | SA7255P | QNN | 35.592 ms | 2 - 9 MB | FP16 | NPU | Use Export Script |
+| Yolo | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 22.79 ms | 0 - 68 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | SA8255 (Proxy) | SA8255P Proxy | QNN | 2.993 ms | 2 - 4 MB | FP16 | NPU | Use Export Script |
+| Yolo | SA8295P ADP | SA8295P | TFLITE | 24.177 ms | 0 - 28 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | SA8295P ADP | SA8295P | QNN | 4.686 ms | 2 - 12 MB | FP16 | NPU | Use Export Script |
+| Yolo | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 22.426 ms | 0 - 68 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | SA8650 (Proxy) | SA8650P Proxy | QNN | 3.01 ms | 2 - 4 MB | FP16 | NPU | Use Export Script |
+| Yolo | SA8775P ADP | SA8775P | TFLITE | 27.588 ms | 0 - 26 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | SA8775P ADP | SA8775P | QNN | 4.599 ms | 2 - 9 MB | FP16 | NPU | Use Export Script |
+| Yolo | QCS8275 (Proxy) | QCS8275 Proxy | TFLITE | 67.992 ms | 0 - 26 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | QCS8275 (Proxy) | QCS8275 Proxy | QNN | 35.592 ms | 2 - 9 MB | FP16 | NPU | Use Export Script |
+| Yolo | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 22.471 ms | 0 - 78 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 2.999 ms | 2 - 5 MB | FP16 | NPU | Use Export Script |
+| Yolo | QCS9075 (Proxy) | QCS9075 Proxy | TFLITE | 27.588 ms | 0 - 26 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | QCS9075 (Proxy) | QCS9075 Proxy | QNN | 4.599 ms | 2 - 9 MB | FP16 | NPU | Use Export Script |
+| Yolo | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 22.488 ms | 0 - 37 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.tflite) |
+| Yolo | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 4.033 ms | 2 - 21 MB | FP16 | NPU | Use Export Script |
+| Yolo | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 3.126 ms | 2 - 2 MB | FP16 | NPU | Use Export Script |
+| Yolo | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 6.548 ms | 10 - 10 MB | FP16 | NPU | [3D-Deep-BOX.onnx](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/Yolo.onnx) |
+| VGG | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 4.776 ms | 0 - 616 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 4.873 ms | 1 - 3 MB | FP16 | NPU | [3D-Deep-BOX.so](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.so) |
+| VGG | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 5.567 ms | 0 - 553 MB | FP16 | NPU | [3D-Deep-BOX.onnx](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.onnx) |
+| VGG | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 3.581 ms | 0 - 36 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 3.842 ms | 0 - 15 MB | FP16 | NPU | [3D-Deep-BOX.so](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.so) |
+| VGG | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 4.37 ms | 1 - 39 MB | FP16 | NPU | [3D-Deep-BOX.onnx](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.onnx) |
+| VGG | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 2.982 ms | 0 - 30 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 3.561 ms | 1 - 31 MB | FP16 | NPU | Use Export Script |
+| VGG | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 4.191 ms | 1 - 33 MB | FP16 | NPU | [3D-Deep-BOX.onnx](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.onnx) |
+| VGG | SA7255P ADP | SA7255P | TFLITE | 257.893 ms | 0 - 24 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | SA7255P ADP | SA7255P | QNN | 258.233 ms | 1 - 8 MB | FP16 | NPU | Use Export Script |
+| VGG | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 4.77 ms | 0 - 612 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | SA8255 (Proxy) | SA8255P Proxy | QNN | 4.865 ms | 1 - 3 MB | FP16 | NPU | Use Export Script |
+| VGG | SA8295P ADP | SA8295P | TFLITE | 9.774 ms | 0 - 26 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | SA8295P ADP | SA8295P | QNN | 9.966 ms | 1 - 11 MB | FP16 | NPU | Use Export Script |
+| VGG | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 4.778 ms | 0 - 612 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | SA8650 (Proxy) | SA8650P Proxy | QNN | 4.865 ms | 1 - 3 MB | FP16 | NPU | Use Export Script |
+| VGG | SA8775P ADP | SA8775P | TFLITE | 10.858 ms | 0 - 24 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | SA8775P ADP | SA8775P | QNN | 11.071 ms | 1 - 8 MB | FP16 | NPU | Use Export Script |
+| VGG | QCS8275 (Proxy) | QCS8275 Proxy | TFLITE | 257.893 ms | 0 - 24 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | QCS8275 (Proxy) | QCS8275 Proxy | QNN | 258.233 ms | 1 - 8 MB | FP16 | NPU | Use Export Script |
+| VGG | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 4.766 ms | 0 - 612 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 4.857 ms | 1 - 4 MB | FP16 | NPU | Use Export Script |
+| VGG | QCS9075 (Proxy) | QCS9075 Proxy | TFLITE | 10.858 ms | 0 - 24 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | QCS9075 (Proxy) | QCS9075 Proxy | QNN | 11.071 ms | 1 - 8 MB | FP16 | NPU | Use Export Script |
+| VGG | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 8.327 ms | 0 - 32 MB | FP16 | NPU | [3D-Deep-BOX.tflite](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.tflite) |
+| VGG | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 8.448 ms | 1 - 33 MB | FP16 | NPU | Use Export Script |
+| VGG | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 5.078 ms | 1 - 1 MB | FP16 | NPU | Use Export Script |
+| VGG | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 5.547 ms | 90 - 90 MB | FP16 | NPU | [3D-Deep-BOX.onnx](https://huggingface.co/qualcomm/3D-Deep-BOX/blob/main/VGG.onnx) |
+## Installation
+Install the package via pip:
+```bash
+pip install "qai-hub-models[deepbox]"
+```
+## Configure Qualcomm® AI Hub to run this model on a cloud-hosted device
+Sign-in to [Qualcomm® AI Hub](https://app.aihub.qualcomm.com/) with your
+Qualcomm® ID. Once signed in navigate to `Account -> Settings -> API Token`.
+With this API token, you can configure your client to run models on the cloud
+hosted devices.
+```bash
+qai-hub configure --api_token API_TOKEN
+```
+Navigate to [docs](https://app.aihub.qualcomm.com/docs/) for more information.
+## Demo off target
+The package contains a simple end-to-end demo that downloads pre-trained
+weights and runs this model on a sample input.
+```bash
+python -m qai_hub_models.models.deepbox.demo
+```
+The above demo runs a reference implementation of pre-processing, model
+inference, and post processing.
+**NOTE**: If you want running in a Jupyter Notebook or Google Colab like
+environment, please add the following to your cell (instead of the above).
+```
+%run -m qai_hub_models.models.deepbox.demo
+```
+### Run model on a cloud-hosted device
+In addition to the demo, you can also run the model on a cloud-hosted Qualcomm®
+device. This script does the following:
+* Performance check on-device on a cloud-hosted device
+* Downloads compiled assets that can be deployed on-device for Android.
+* Accuracy check between PyTorch and on-device outputs.
+```bash
+python -m qai_hub_models.models.deepbox.export
+```
+```
+Profiling Results
+------------------------------------------------------------
+Yolo
+Device                          : Samsung Galaxy S23 (13)
+Runtime                         : TFLITE
+Estimated inference time (ms)   : 22.2
+Estimated peak memory usage (MB): [0, 59]
+Total # Ops                     : 128
+Compute Unit(s)                 : NPU (128 ops)
+------------------------------------------------------------
+VGG
+Device                          : Samsung Galaxy S23 (13)
+Runtime                         : TFLITE
+Estimated inference time (ms)   : 4.8
+Estimated peak memory usage (MB): [0, 616]
+Total # Ops                     : 40
+Compute Unit(s)                 : NPU (40 ops)
+```
+## How does this work?
+This [export script](https://aihub.qualcomm.com/models/deepbox/qai_hub_models/models/3D-Deep-BOX/export.py)
+leverages [Qualcomm® AI Hub](https://aihub.qualcomm.com/) to optimize, validate, and deploy this model
+on-device. Lets go through each step below in detail:
+Step 1: **Compile model for on-device deployment**
+To compile a PyTorch model for on-device deployment, we first trace the model
+in memory using the `jit.trace` and then call the `submit_compile_job` API.
+```python
+import torch
+import qai_hub as hub
+from qai_hub_models.models.deepbox import Model
+# Load the model
+model = Model.from_pretrained()
+bbox2D_dectector_model = model.bbox2D_dectector
+bbox3D_dectector_model = model.bbox3D_dectector
+# Device
+device = hub.Device("Samsung Galaxy S23")
+# Trace model
+bbox2D_dectector_input_shape = bbox2D_dectector_model.get_input_spec()
+bbox2D_dectector_sample_inputs = bbox2D_dectector_model.sample_inputs()
+traced_bbox2D_dectector_model = torch.jit.trace(bbox2D_dectector_model, [torch.tensor(data[0]) for _, data in bbox2D_dectector_sample_inputs.items()])
+# Compile model on a specific device
+bbox2D_dectector_compile_job = hub.submit_compile_job(
+    model=traced_bbox2D_dectector_model ,
+    device=device,
+    input_specs=bbox2D_dectector_model.get_input_spec(),
+)
+# Get target model to run on-device
+bbox2D_dectector_target_model = bbox2D_dectector_compile_job.get_target_model()
+# Trace model
+bbox3D_dectector_input_shape = bbox3D_dectector_model.get_input_spec()
+bbox3D_dectector_sample_inputs = bbox3D_dectector_model.sample_inputs()
+traced_bbox3D_dectector_model = torch.jit.trace(bbox3D_dectector_model, [torch.tensor(data[0]) for _, data in bbox3D_dectector_sample_inputs.items()])
+# Compile model on a specific device
+bbox3D_dectector_compile_job = hub.submit_compile_job(
+    model=traced_bbox3D_dectector_model ,
+    device=device,
+    input_specs=bbox3D_dectector_model.get_input_spec(),
+)
+# Get target model to run on-device
+bbox3D_dectector_target_model = bbox3D_dectector_compile_job.get_target_model()
+```
+Step 2: **Performance profiling on cloud-hosted device**
+After compiling models from step 1. Models can be profiled model on-device using the
+`target_model`. Note that this scripts runs the model on a device automatically
+provisioned in the cloud.  Once the job is submitted, you can navigate to a
+provided job URL to view a variety of on-device performance metrics.
+```python
+bbox2D_dectector_profile_job = hub.submit_profile_job(
+    model=bbox2D_dectector_target_model,
+    device=device,
+)
+bbox3D_dectector_profile_job = hub.submit_profile_job(
+    model=bbox3D_dectector_target_model,
+    device=device,
+)
+```
+Step 3: **Verify on-device accuracy**
+To verify the accuracy of the model on-device, you can run on-device inference
+on sample input data on the same cloud hosted device.
+```python
+bbox2D_dectector_input_data = bbox2D_dectector_model.sample_inputs()
+bbox2D_dectector_inference_job = hub.submit_inference_job(
+    model=bbox2D_dectector_target_model,
+    device=device,
+    inputs=bbox2D_dectector_input_data,
+)
+bbox2D_dectector_inference_job.download_output_data()
+bbox3D_dectector_input_data = bbox3D_dectector_model.sample_inputs()
+bbox3D_dectector_inference_job = hub.submit_inference_job(
+    model=bbox3D_dectector_target_model,
+    device=device,
+    inputs=bbox3D_dectector_input_data,
+)
+bbox3D_dectector_inference_job.download_output_data()
+```
+With the output of the model, you can compute like PSNR, relative errors or
+spot check the output with expected output.
+**Note**: This on-device profiling and inference requires access to Qualcomm®
+AI Hub. [Sign up for access](https://myaccount.qualcomm.com/signup).
+## Deploying compiled model to Android
+The models can be deployed using multiple runtimes:
+- TensorFlow Lite (`.tflite` export): [This
+  tutorial](https://www.tensorflow.org/lite/android/quickstart) provides a
+  guide to deploy the .tflite model in an Android application.
+- QNN (`.so` export ): This [sample
+  app](https://docs.qualcomm.com/bundle/publicresource/topics/80-63442-50/sample_app.html)
+provides instructions on how to use the `.so` shared library  in an Android application.
+## View on Qualcomm® AI Hub
+Get more details on 3D-Deep-BOX's performance across various devices [here](https://aihub.qualcomm.com/models/deepbox).
+Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/)
+## License
+* The license for the original implementation of 3D-Deep-BOX can be found
+  [here](https://github.com/skhadem/3D-BoundingBox/blob/master/LICENSE).
+* The license for the compiled assets for on-device deployment can be found [here](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/Qualcomm+AI+Hub+Proprietary+License.pdf)
+## References
+* [3D Bounding Box Estimation Using Deep Learning and Geometry](https://arxiv.org/abs/1612.00496)
+* [Source Model Implementation](https://github.com/skhadem/3D-BoundingBox/)
+## 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:[email protected]).