Hugging Face's logo

qualcomm
/
TrOCR

Image-to-Text
PyTorch
TF Lite
ONNX
android
Model card Files Community
2
TrOCR / README.md
qaihm-bot's picture
qaihm-bot
Upload README.md with huggingface_hub
5d5f5af verified
preview code
|
raw
history blame
16.5 kB
---
library_name: pytorch
license: mit
pipeline_tag: image-to-text
tags:
- android
---
![](https://qaihub-public-assets.s3.us-west-2.amazonaws.com/qai-hub-models/models/trocr/web-assets/model_demo.png)
# 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
|---|---|---|---|---|---|---|---|---|
| TrOCRDecoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 2.172 ms | 0 - 303 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
| TrOCRDecoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 2.376 ms | 0 - 270 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.so) |
| TrOCRDecoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 3.019 ms | 0 - 246 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) |
| TrOCRDecoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 1.58 ms | 0 - 52 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
| TrOCRDecoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 1.836 ms | 0 - 52 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.so) |
| TrOCRDecoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 2.074 ms | 0 - 61 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) |
| TrOCRDecoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 1.45 ms | 0 - 47 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
| TrOCRDecoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 1.782 ms | 0 - 47 MB | FP16 | NPU | Use Export Script |
| TrOCRDecoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 2.123 ms | 0 - 45 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) |
| TrOCRDecoder | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 2.187 ms | 0 - 289 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
| TrOCRDecoder | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 2.265 ms | 2 - 4 MB | FP16 | NPU | Use Export Script |
| TrOCRDecoder | SA7255P ADP | SA7255P | TFLITE | 12.254 ms | 0 - 43 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
| TrOCRDecoder | SA7255P ADP | SA7255P | QNN | 12.375 ms | 7 - 16 MB | FP16 | NPU | Use Export Script |
| TrOCRDecoder | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 2.207 ms | 0 - 272 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
| TrOCRDecoder | SA8255 (Proxy) | SA8255P Proxy | QNN | 2.263 ms | 1 - 4 MB | FP16 | NPU | Use Export Script |
| TrOCRDecoder | SA8295P ADP | SA8295P | TFLITE | 3.11 ms | 0 - 44 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
| TrOCRDecoder | SA8295P ADP | SA8295P | QNN | 4.001 ms | 7 - 21 MB | FP16 | NPU | Use Export Script |
| TrOCRDecoder | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 2.184 ms | 0 - 372 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
| TrOCRDecoder | SA8650 (Proxy) | SA8650P Proxy | QNN | 2.272 ms | 2 - 5 MB | FP16 | NPU | Use Export Script |
| TrOCRDecoder | SA8775P ADP | SA8775P | TFLITE | 3.339 ms | 0 - 44 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
| TrOCRDecoder | SA8775P ADP | SA8775P | QNN | 3.525 ms | 7 - 17 MB | FP16 | NPU | Use Export Script |
| TrOCRDecoder | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 2.505 ms | 0 - 48 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.tflite) |
| TrOCRDecoder | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 2.756 ms | 4 - 54 MB | FP16 | NPU | Use Export Script |
| TrOCRDecoder | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 2.42 ms | 7 - 7 MB | FP16 | NPU | Use Export Script |
| TrOCRDecoder | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 2.757 ms | 69 - 69 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCRDecoder.onnx) |
| TrOCREncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | TFLITE | 50.082 ms | 7 - 30 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
| TrOCREncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | QNN | 52.43 ms | 2 - 19 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.so) |
| TrOCREncoder | Samsung Galaxy S23 | Snapdragon® 8 Gen 2 | ONNX | 39.313 ms | 14 - 157 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.onnx) |
| TrOCREncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | TFLITE | 38.871 ms | 5 - 68 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
| TrOCREncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | QNN | 40.61 ms | 2 - 63 MB | FP16 | NPU | [TrOCR.so](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.so) |
| TrOCREncoder | Samsung Galaxy S24 | Snapdragon® 8 Gen 3 | ONNX | 31.086 ms | 14 - 73 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.onnx) |
| TrOCREncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | TFLITE | 36.23 ms | 7 - 71 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
| TrOCREncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | QNN | 33.584 ms | 2 - 66 MB | FP16 | NPU | Use Export Script |
| TrOCREncoder | Snapdragon 8 Elite QRD | Snapdragon® 8 Elite | ONNX | 26.529 ms | 16 - 78 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.onnx) |
| TrOCREncoder | QCS8550 (Proxy) | QCS8550 Proxy | TFLITE | 49.996 ms | 7 - 31 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
| TrOCREncoder | QCS8550 (Proxy) | QCS8550 Proxy | QNN | 37.253 ms | 2 - 4 MB | FP16 | NPU | Use Export Script |
| TrOCREncoder | SA7255P ADP | SA7255P | TFLITE | 266.112 ms | 1 - 63 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
| TrOCREncoder | SA7255P ADP | SA7255P | QNN | 247.638 ms | 2 - 11 MB | FP16 | NPU | Use Export Script |
| TrOCREncoder | SA8255 (Proxy) | SA8255P Proxy | TFLITE | 50.345 ms | 7 - 30 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
| TrOCREncoder | SA8255 (Proxy) | SA8255P Proxy | QNN | 37.553 ms | 2 - 4 MB | FP16 | NPU | Use Export Script |
| TrOCREncoder | SA8295P ADP | SA8295P | TFLITE | 65.333 ms | 7 - 68 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
| TrOCREncoder | SA8295P ADP | SA8295P | QNN | 50.544 ms | 2 - 16 MB | FP16 | NPU | Use Export Script |
| TrOCREncoder | SA8650 (Proxy) | SA8650P Proxy | TFLITE | 50.38 ms | 7 - 29 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
| TrOCREncoder | SA8650 (Proxy) | SA8650P Proxy | QNN | 37.52 ms | 2 - 4 MB | FP16 | NPU | Use Export Script |
| TrOCREncoder | SA8775P ADP | SA8775P | TFLITE | 59.748 ms | 7 - 69 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
| TrOCREncoder | SA8775P ADP | SA8775P | QNN | 42.265 ms | 2 - 12 MB | FP16 | NPU | Use Export Script |
| TrOCREncoder | QCS8450 (Proxy) | QCS8450 Proxy | TFLITE | 60.415 ms | 7 - 66 MB | FP16 | NPU | [TrOCR.tflite](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.tflite) |
| TrOCREncoder | QCS8450 (Proxy) | QCS8450 Proxy | QNN | 64.846 ms | 0 - 63 MB | FP16 | NPU | Use Export Script |
| TrOCREncoder | Snapdragon X Elite CRD | Snapdragon® X Elite | QNN | 34.064 ms | 2 - 2 MB | FP16 | NPU | Use Export Script |
| TrOCREncoder | Snapdragon X Elite CRD | Snapdragon® X Elite | ONNX | 36.717 ms | 49 - 49 MB | FP16 | NPU | [TrOCR.onnx](https://huggingface.co/qualcomm/TrOCR/blob/main/TrOCREncoder.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
------------------------------------------------------------
TrOCRDecoder
Device : Samsung Galaxy S23 (13)
Runtime : TFLITE
Estimated inference time (ms) : 2.2
Estimated peak memory usage (MB): [0, 303]
Total # Ops : 399
Compute Unit(s) : NPU (399 ops)
------------------------------------------------------------
TrOCREncoder
Device : Samsung Galaxy S23 (13)
Runtime : TFLITE
Estimated inference time (ms) : 50.1
Estimated peak memory usage (MB): [7, 30]
Total # Ops : 591
Compute Unit(s) : NPU (591 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 Model
# Load the model
model = Model.from_pretrained()
decoder_model = model.decoder
encoder_model = model.encoder
# Device
device = hub.Device("Samsung Galaxy S23")
# 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()
# 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()
```
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
decoder_profile_job = hub.submit_profile_job(
model=decoder_target_model,
device=device,
)
encoder_profile_job = hub.submit_profile_job(
model=encoder_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
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()
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()
```
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:ai-hub-support@qti.qualcomm.com).