AsterOCR 🇻🇳
AsterOCR is a state-of-the-art, ultra-fast neural Vietnamese Optical Character Recognition (OCR) engine. Designed as a modern, non-autoregressive replacement for legacy OCR systems like VietOCR, AsterOCR provides sub-60ms recognition speeds and exceptional accuracy across administrative records (CCCD, Passports, Driver's Licenses), invoices, financial receipts, and degraded smartphone captures.
⚡ Quickstart (Python)
1. Installation
pip install asterocr
(No heavy PyTorch installation required! Uses high-performance ONNX Runtime by default.)
2. Python API
from asterocr import AsterOCR
# Initialize engine (automatically downloads weights on first run)
ocr = AsterOCR()
# Recognize text from file path, PIL Image, or bytes
text = ocr.predict("path/to/invoice.jpg")
print(text)
# Output: CỘNG HÒA XÃ HỘI CHỦ NGHĨA VIỆT NAM
3. Command Line Interface (CLI)
asterocr path/to/image.jpg
🌐 Cross-Platform Distribution Ecosystem
AsterOCR models are exported in universal ONNX format (~40MB) hosted on Hugging Face Hub:
| Platform | Package | Engine | Status |
|---|---|---|---|
| Python | asterocr |
onnxruntime / torch |
✅ Ready (PyPI) |
| Node.js | @asterxlabs/asterocr |
onnxruntime-node |
✅ Ready (npm) |
| Flutter | asterocr_flutter |
onnxruntime_flutter |
✅ Ready (pub.dev) |
| Hugging Face | asterxlabs/asternet-vietnamese-ocr |
Model Hub & Space Demo | ✅ Active |
📊 Benchmarks vs Industry Standard
| Metric | VietOCR (Transformer) | AsterOCR v1.0 | Improvement |
|---|---|---|---|
| Exact Match (EM) | 71.4% | 86.0% | +14.6% |
| Character Error Rate (CER) | 7.82% | 3.45% | -55.8% |
| CPU Inference Latency | 340ms | 60.2ms | 5.6x faster |
| Model Size | 145 MB | 40.76 MB | 72% smaller |
📜 License
MIT License © 2026 asterxlabs.
1 maintainer
Metadata
1 maintainer
Release files for asterocr 1.0.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | |
|---|---|---|---|---|
| asterocr-1.0.0.tar.gz | 117.5 kB | Without provenance | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| asterocr-1.0.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 249.1 kB
Release files / asterocr-1.0.0.tar.gz
| Download URL | asterocr-1.0.0.tar.gz |
|---|---|
| Size | 117.5 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
37a17526e74a5b32b1febd5f8985f4bd252d24052ac4052a0e95c152c830b38a
|
|
BLAKE2b-256 checksum How to use checksums |
0fe1cbb338eaf84e0e6aaaa124dfa29c05aab83102250febb5c6e1c89e0485d6
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.12.10
|
Release files / asterocr-1.0.0-py3-none-any.whl
| Download URL | asterocr-1.0.0-py3-none-any.whl |
|---|---|
| Size | 131.6 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
8ef3a744c3f0f55a039591c49e3992645eb5d32684bd4f24ccbf567f5dc13791
|
|
BLAKE2b-256 checksum How to use checksums |
79bddfbfd478324411b46e1bfbf04d64c3c18d09c570320006425a24155ea2a1
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.12.10
|
Security for asterocr 1.0.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.
