|
--- |
|
library_name: pytorch |
|
license: mit |
|
pipeline_tag: image-to-text |
|
tags: |
|
- android |
|
|
|
--- |
|
|
|
 |
|
|
|
# TrOCR: Optimized for Mobile Deployment |
|
## Transformer based model for state-of-the-art optical character recognition (OCR) on both printed and handwritten text |
|
|
|
|
|
End-to-end text recognition approach with pre-trained image transformer and text transformer models for both image understanding and wordpiece-level text generation. |
|
|
|
This model is an implementation of TrOCR found [here](https://huggingface.co/microsoft/trocr-small-stage1). |
|
|
|
|
|
This repository provides scripts to run TrOCR on Qualcomm® devices. |
|
More details on model performance across various devices, can be found |
|
[here](https://aihub.qualcomm.com/models/trocr). |
|
|
|
|
|
### Model Details |
|
|
|
- **Model Type:** Image to text |
|
- **Model Stats:** |
|
- Model checkpoint: trocr-small-stage1 |
|
- Input resolution: 320x320 |
|
- Number of parameters (TrOCREncoder): 23.0M |
|
- Model size (TrOCREncoder): 87.8 MB |
|
- Number of parameters (TrOCRDecoder): 38.3M |
|
- Model size (TrOCRDecoder): 146 MB |
|
|
|
| Model | Device | Chipset | Target Runtime | Inference Time (ms) | Peak Memory Range (MB) | Precision | Primary Compute Unit | Target Model |
|
|---|---|---|---|---|---|---|---|---| |
|
| TrOCREncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 50.57 ms | 7 - 9 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) | |
|
| TrOCREncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 51.644 ms | 2 - 23 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.so) | |
|
| TrOCREncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 38.248 ms | 0 - 112 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.onnx) | |
|
| TrOCREncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 40.227 ms | 6 - 311 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) | |
|
| TrOCREncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 41.582 ms | 2 - 66 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.so) | |
|
| TrOCREncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 29.203 ms | 0 - 346 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.onnx) | |
|
| TrOCREncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 32.548 ms | 5 - 122 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) | |
|
| TrOCREncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 33.111 ms | 2 - 66 MB | FP16 | NPU | Use Export Script | |
|
| TrOCREncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 25.501 ms | 16 - 218 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.onnx) | |
|
| TrOCREncoder | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 50.102 ms | 7 - 9 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) | |
|
| TrOCREncoder | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 36.279 ms | 2 - 3 MB | FP16 | NPU | Use Export Script | |
|
| TrOCREncoder | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 50.44 ms | 7 - 9 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) | |
|
| TrOCREncoder | SA8255 (Proxy) | SA8255P Proxy | QNN | 36.944 ms | 2 - 4 MB | FP16 | NPU | Use Export Script | |
|
| TrOCREncoder | SA8775 (Proxy) | SA8775P Proxy | TFLITE | 51.532 ms | 7 - 9 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) | |
|
| TrOCREncoder | SA8775 (Proxy) | SA8775P Proxy | QNN | 36.72 ms | 2 - 3 MB | FP16 | NPU | Use Export Script | |
|
| TrOCREncoder | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 50.666 ms | 7 - 9 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) | |
|
| TrOCREncoder | SA8650 (Proxy) | SA8650P Proxy | QNN | 36.616 ms | 2 - 4 MB | FP16 | NPU | Use Export Script | |
|
| TrOCREncoder | SA8295P ADP | SA8295P | TFLITE | 65.458 ms | 7 - 113 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) | |
|
| TrOCREncoder | SA8295P ADP | SA8295P | QNN | 50.099 ms | 2 - 7 MB | FP16 | NPU | Use Export Script | |
|
| TrOCREncoder | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 61.165 ms | 7 - 301 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) | |
|
| TrOCREncoder | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 60.141 ms | 2 - 67 MB | FP16 | NPU | Use Export Script | |
|
| TrOCREncoder | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 33.869 ms | 2 - 2 MB | FP16 | NPU | Use Export Script | |
|
| TrOCREncoder | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 35.826 ms | 109 - 109 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.onnx) | |
|
| TrOCRDecoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 2.631 ms | 0 - 869 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) | |
|
| TrOCRDecoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 2.924 ms | 0 - 126 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.so) | |
|
| TrOCRDecoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 2.924 ms | 0 - 74 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) | |
|
| TrOCRDecoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 1.848 ms | 0 - 189 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) | |
|
| TrOCRDecoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 2.103 ms | 0 - 50 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.so) | |
|
| TrOCRDecoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 1.991 ms | 0 - 148 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) | |
|
| TrOCRDecoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.961 ms | 0 - 27 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) | |
|
| TrOCRDecoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 1.744 ms | 0 - 45 MB | FP16 | NPU | Use Export Script | |
|
| TrOCRDecoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 1.788 ms | 0 - 34 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) | |
|
| TrOCRDecoder | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 2.551 ms | 0 - 993 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) | |
|
| TrOCRDecoder | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 2.629 ms | 1 - 2 MB | FP16 | NPU | Use Export Script | |
|
| TrOCRDecoder | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 2.626 ms | 0 - 2 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) | |
|
| TrOCRDecoder | SA8255 (Proxy) | SA8255P Proxy | QNN | 2.671 ms | 2 - 4 MB | FP16 | NPU | Use Export Script | |
|
| TrOCRDecoder | SA8775 (Proxy) | SA8775P Proxy | TFLITE | 2.585 ms | 0 - 2 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) | |
|
| TrOCRDecoder | SA8775 (Proxy) | SA8775P Proxy | QNN | 2.676 ms | 1 - 3 MB | FP16 | NPU | Use Export Script | |
|
| TrOCRDecoder | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 2.625 ms | 0 - 2 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) | |
|
| TrOCRDecoder | SA8650 (Proxy) | SA8650P Proxy | QNN | 2.661 ms | 0 - 2 MB | FP16 | NPU | Use Export Script | |
|
| TrOCRDecoder | SA8295P ADP | SA8295P | TFLITE | 3.318 ms | 0 - 25 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) | |
|
| TrOCRDecoder | SA8295P ADP | SA8295P | QNN | 3.953 ms | 7 - 12 MB | FP16 | NPU | Use Export Script | |
|
| TrOCRDecoder | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 2.868 ms | 0 - 189 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) | |
|
| TrOCRDecoder | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 3.325 ms | 6 - 51 MB | FP16 | NPU | Use Export Script | |
|
| TrOCRDecoder | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 2.84 ms | 7 - 7 MB | FP16 | NPU | Use Export Script | |
|
| TrOCRDecoder | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 2.887 ms | 68 - 68 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) | |
|
|
|
|
|
|
|
|
|
## Installation |
|
|
|
This model can be installed as a Python package via pip. |
|
|
|
```bash |
|
pip install "qai-hub-models[trocr]" |
|
``` |
|
|
|
|
|
|
|
## 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.trocr.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.trocr.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.trocr.export |
|
``` |
|
``` |
|
Profiling Results |
|
------------------------------------------------------------ |
|
TrOCREncoder |
|
Device : Samsung Galaxy S23 (13) |
|
Runtime : TFLITE |
|
Estimated inference time (ms) : 50.6 |
|
Estimated peak memory usage (MB): [7, 9] |
|
Total # Ops : 591 |
|
Compute Unit(s) : NPU (591 ops) |
|
|
|
------------------------------------------------------------ |
|
TrOCRDecoder |
|
Device : Samsung Galaxy S23 (13) |
|
Runtime : TFLITE |
|
Estimated inference time (ms) : 2.6 |
|
Estimated peak memory usage (MB): [0, 869] |
|
Total # Ops : 399 |
|
Compute Unit(s) : NPU (399 ops) |
|
``` |
|
|
|
|
|
## How does this work? |
|
|
|
This [export script](https://aihub.qualcomm.com/models/trocr/qai_hub_models/models/TrOCR/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.trocr import TrOCREncoder,TrOCRDecoder |
|
|
|
# Load the model |
|
encoder_model = TrOCREncoder.from_pretrained() |
|
decoder_model = TrOCRDecoder.from_pretrained() |
|
|
|
# Device |
|
device = hub.Device("Samsung Galaxy S23") |
|
|
|
# Trace model |
|
encoder_input_shape = encoder_model.get_input_spec() |
|
encoder_sample_inputs = encoder_model.sample_inputs() |
|
|
|
traced_encoder_model = torch.jit.trace(encoder_model, [torch.tensor(data[0]) for _, data in encoder_sample_inputs.items()]) |
|
|
|
# Compile model on a specific device |
|
encoder_compile_job = hub.submit_compile_job( |
|
model=traced_encoder_model , |
|
device=device, |
|
input_specs=encoder_model.get_input_spec(), |
|
) |
|
|
|
# Get target model to run on-device |
|
encoder_target_model = encoder_compile_job.get_target_model() |
|
# Trace model |
|
decoder_input_shape = decoder_model.get_input_spec() |
|
decoder_sample_inputs = decoder_model.sample_inputs() |
|
|
|
traced_decoder_model = torch.jit.trace(decoder_model, [torch.tensor(data[0]) for _, data in decoder_sample_inputs.items()]) |
|
|
|
# Compile model on a specific device |
|
decoder_compile_job = hub.submit_compile_job( |
|
model=traced_decoder_model , |
|
device=device, |
|
input_specs=decoder_model.get_input_spec(), |
|
) |
|
|
|
# Get target model to run on-device |
|
decoder_target_model = decoder_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 |
|
encoder_profile_job = hub.submit_profile_job( |
|
model=encoder_target_model, |
|
device=device, |
|
) |
|
decoder_profile_job = hub.submit_profile_job( |
|
model=decoder_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 |
|
encoder_input_data = encoder_model.sample_inputs() |
|
encoder_inference_job = hub.submit_inference_job( |
|
model=encoder_target_model, |
|
device=device, |
|
inputs=encoder_input_data, |
|
) |
|
encoder_inference_job.download_output_data() |
|
decoder_input_data = decoder_model.sample_inputs() |
|
decoder_inference_job = hub.submit_inference_job( |
|
model=decoder_target_model, |
|
device=device, |
|
inputs=decoder_input_data, |
|
) |
|
decoder_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 TrOCR's performance across various devices [here](https://aihub.qualcomm.com/models/trocr). |
|
Explore all available models on [Qualcomm® AI Hub](https://aihub.qualcomm.com/) |
|
|
|
|
|
## License |
|
* The license for the original implementation of TrOCR can be found [here](https://github.com/microsoft/unilm/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 |
|
* [TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models](https://arxiv.org/abs/2109.10282) |
|
* [Source Model Implementation](https://huggingface.co/microsoft/trocr-small-stage1) |
|
|
|
|
|
|
|
## 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]). |
|
|
|
|
|
|
