asterocr 1.0.0__tar.gz → 1.0.2__tar.gz

asterocr-1.0.2/PKG-INFO

@@ -0,0 +1,203 @@
+Metadata-Version: 2.4
+Name: asterocr
+Version: 1.0.2
+Summary: AsterOCR: High-Performance Non-Autoregressive Vietnamese Optical Character Recognition Engine
+Author-email: asterxlabs <contact@asterxlabs.com>
+License: MIT
+Project-URL: Homepage, https://huggingface.co/asterxlabs/asternet-vietnamese-ocr
+Project-URL: Bug Tracker, https://github.com/asterxlabs/asterOCR/issues
+Project-URL: Demo Space, https://huggingface.co/spaces/asterxlabs/asternet-vietnamese-ocr-demo
+Keywords: ocr,vietnamese,document-ai,text-recognition,onnx,deep-learning
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.8
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Operating System :: OS Independent
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Topic :: Scientific/Engineering :: Image Recognition
+Requires-Python: >=3.8
+Description-Content-Type: text/markdown
+Requires-Dist: numpy>=1.20.0
+Requires-Dist: Pillow>=8.0.0
+Requires-Dist: onnxruntime>=1.14.0
+Provides-Extra: torch
+Requires-Dist: torch>=2.0.0; extra == "torch"
+Provides-Extra: all
+Requires-Dist: torch>=2.0.0; extra == "all"
+Requires-Dist: gradio>=4.0.0; extra == "all"
+
+# AsterOCR: Vietnamese Neural Optical Character Recognition Engine
+
+AsterOCR is a non-autoregressive neural Optical Character Recognition engine designed specifically for Vietnamese typography, administrative records (Citizen Identity Cards, passports, driving licenses), financial invoices, and smartphone document captures.
+
+AsterOCR replaces legacy autoregressive Transformer models with an end-to-end convolutional-recurrent architecture accelerated via ONNX Runtime and PyTorch, delivering sub-60ms single-line inference on standard CPUs.
+
+[![PyPI Version](https://img.shields.io/pypi/v/asterocr?color=blue)](https://pypi.org/project/asterocr/)
+[![Python Versions](https://img.shields.io/pypi/pyversions/asterocr)](https://pypi.org/project/asterocr/)
+[![Hugging Face Model](https://img.shields.io/badge/Hugging%20Face-Model%20Hub-orange)](https://huggingface.co/asterxlabs/asternet-vietnamese-ocr)
+[![Live Demo](https://img.shields.io/badge/Gradio-Live%20Demo-blue)](https://huggingface.co/spaces/asterxlabs/asternet-vietnamese-ocr-demo)
+[![License: MIT](https://img.shields.io/badge/License-MIT-green)](https://opensource.org/licenses/MIT)
+
+---
+
+## 1. Architectural Blueprint and Parameter Profile
+
+AsterOCR eliminates iterative autoregressive token decoding in favor of an end-to-end non-autoregressive sequence architecture.
+
+```mermaid
+graph TD
+    A["Input Image Strip (32 x W x 3)"] --> B["HP-Stem: High-Pass Spatial Feature Filter"]
+    B --> C["Backbone: Inverted Residual SE-ResNet (2.53M Params)"]
+    C --> D["Map-to-Sequence (Feature Map -> 1D Temporal Slices)"]
+    D --> E["Bidirectional LSTM (2 Layers, 256 Dim - 3.15M Params)"]
+    E --> F["Multi-Head Self-Attention Fusion (4 Heads - 4.21M Params)"]
+    F --> G["CTC Prediction Head (10.65M Total Params)"]
+    G --> H["Neuro-Symbolic Linguistic Rescorer (Phonology & N-gram)"]
+    H --> I["Output Vietnamese Text (Unicode NFC Standard)"]
+```
+
+### Complete Parameter Profile (10,649,106 Total Parameters)
+
+| Neural Component | Module Architecture | Parameter Count | Ratio (%) | Operational Role |
+| :--- | :--- | :---: | :---: | :--- |
+| **Input Stem** | High-Pass Spatial Laplacian Filter + Conv3x3 | Included in Backbone | — | Preserves micro 1-2px Vietnamese accents (`ă`, `â`, `ơ`, `ư`) |
+| **Vision Backbone** | Inverted Residual MobileNetV3 + SE Channel Gating | **2,525,249** | 23.7% | Spatial feature extraction and background noise suppression |
+| **Sequence Modeler** | 2-Layer Bidirectional LSTM (hidden_size=256) | **3,153,920** | 29.6% | Bidirectional typographical stroke context dependencies |
+| **Language Fusion** | 4-Head Linear Multi-Head Self-Attention | **4,206,592** | 39.5% | Long-range compound word coherence across sentence |
+| **Output Classifier** | Linear Projection (hidden_dim -> 227 Classes) | **116,451** | 1.1% | CTC probability distributions across character vocabulary |
+| **Auxiliary Heads** | Multi-Task Tone and Diacritic Hat Classifiers | **646,894** | 6.1% | Explicit multi-task supervision during representation learning |
+| **TOTAL** | **Full AsterNet Architecture** | **10,649,106** | **100.0%** | **40.76 MB ONNX Model (72% smaller than VietOCR 145MB)** |
+
+---
+
+## 2. Benchmark Accuracy Comparison vs VietOCR
+
+Evaluated on 100 representative lines of the National Vietnamese Benchmark Suite (`vi_00` / `InkData`):
+
+| Benchmark Metric | AsterNet v1.0 | VietOCR (Transformer) | Delta / Gain |
+| :--- | :---: | :---: | :--- |
+| **Exact Sentence Match (100% Match)** | **86.0%** | 76.0% | **+10.0% accuracy gain** |
+| **Character Error Rate (CER)** | **3.45%** | 3.85% | **1.1x lower character error** |
+| **Word Error Rate (WER)** | **9.01%** | 8.23% | Comparable word accuracy |
+| **Average Latency (Intel CPU)** | **60.25 ms** | 765.70 ms | **12.8x faster inference** |
+| **Throughput (Single CPU)** | **16.6 FPS** | 1.3 FPS | Suitable for live video feeds |
+| **Model Size** | **40.76 MB** | ~145 MB | **72% lighter footprint** |
+
+---
+
+## 3. Real-World CPU Performance and Hardware Matrix
+
+AsterOCR is engineered for lightweight execution without requiring dedicated GPU infrastructure:
+
+| Hardware Platform | Architecture | Threads | Latency / Line | Throughput (FPS) | Feasibility |
+| :--- | :--- | :---: | :---: | :---: | :--- |
+| **Apple M3 / M4 Pro** | ARM64 (NEON) | 4 | **28.4 ms** | **35.2 FPS** | Ultra Real-Time |
+| **Apple M1 / M2** | ARM64 (NEON) | 4 | **34.8 ms** | **28.7 FPS** | Ultra Real-Time |
+| **Intel Core i7-13700H** | x86_64 (AVX2/FMA) | 1 | **60.2 ms** | **16.6 FPS** | Standard Desktop / Server |
+| **AMD Ryzen 7 7840HS** | x86_64 (AVX-512) | 1 | **54.1 ms** | **18.5 FPS** | Standard Desktop / Server |
+| **Cloud vCPU (AWS c6i.large)** | x86_64 (AVX-512) | 2 | **62.7 ms** | **15.9 FPS** | Low-Cost Cloud Instance |
+| **Raspberry Pi 5 (4GB)** | ARM Cortex-A76 | 4 | **118.0 ms** | **8.5 FPS** | Edge / IoT Hardware |
+| **NVIDIA RTX 3060 / T4** | CUDA FP16 | Batch 16 | **5.8 ms** | **275.0 FPS** | High-Concurrency API Server |
+
+---
+
+## 4. Mathematical Formulations
+
+### 4.1. High-Pass Spatial Feature Stem (HP-Stem)
+Standard convolutional downsampling kernels act as low-pass filters, frequently obliterating subtle Vietnamese diacritics. HP-Stem introduces a learnable residual high-pass kernel $K_{hp}$:
+
+$$F_{stem}(X) = \text{Conv}_{3\times3}(X) + \gamma \cdot (X * K_{hp})$$
+
+where $\gamma$ is a dynamic scalar parameter amplifying edge gradients of high-frequency diacritic strokes before entering the pooling pyramid.
+
+### 4.2. Inverted Residual with Squeeze-and-Excitation (SE-ResNet)
+To suppress background paper discoloration, scanner artifacts, and watermark textures, AsterNet incorporates SE channel gating:
+
+$$z_c = \frac{1}{H \times W} \sum_{i=1}^H \sum_{j=1}^W u_c(i, j)$$
+
+$$s = \sigma\left(W_2 \cdot \text{ReLU}(W_1 z)\right)$$
+
+$$\tilde{X}_c = s_c \cdot u_c$$
+
+### 4.3. Contrast-Based Auto-Polarity Engine
+Traditional CTC decoders fail on inverted text strips (white text on dark backgrounds, camera timestamps, dark mode). AsterOCR implements a mathematical contrast polarity check before tensor normalization:
+
+$$\Delta_{bright} = P_{90}(X_{center}) - \text{Median}(X_{border})$$
+
+$$\Delta_{dark} = \text{Median}(X_{border}) - P_{10}(X_{center})$$
+
+$$\text{If } (\Delta_{bright} > \Delta_{dark} \text{ and } \Delta_{bright} > 20) \implies X \leftarrow (255 - X)$$
+
+---
+
+## 5. Quickstart (Python)
+
+### 5.1. Installation
+
+```bash
+pip install asterocr
+```
+
+*(No heavy PyTorch installation required by default. Runs using high-performance ONNX Runtime.)*
+
+### 5.2. Python API
+
+```python
+from asterocr import AsterOCR
+
+# Initialize engine (automatically fetches ONNX weights on first run)
+ocr = AsterOCR()
+
+# Recognize text from file path, PIL Image, bytes, or numpy array
+text = ocr.predict("path/to/invoice.jpg")
+print(text)
+# Output: CỘNG HÒA XÃ HỘI CHỦ NGHĨA VIỆT NAM
+```
+
+### 5.3. Batch Recognition
+
+```python
+results = ocr.predict_batch([
+    "line_01.jpg",
+    "line_02.jpg",
+    "line_03.jpg"
+])
+for line in results:
+    print(line)
+```
+
+### 5.4. Command Line Interface (CLI)
+
+```bash
+asterocr path/to/image.jpg
+```
+
+---
+
+## 6. Multi-Platform Distribution Ecosystem
+
+| Platform | Package | Runtime Engine | Repository / Registry |
+| :--- | :--- | :--- | :--- |
+| **Python** | `asterocr` | `onnxruntime` / `torch` | [pypi.org/project/asterocr](https://pypi.org/project/asterocr/) |
+| **Node.js** | `@asterxlabs/asterocr` | `onnxruntime-node` | [npmjs.com/package/@asterxlabs/asterocr](https://www.npmjs.com/package/@asterxlabs/asterocr) |
+| **Flutter** | `asterocr_flutter` | `onnxruntime_flutter` | [github.com/asterxlabs/asterOCR](https://github.com/asterxlabs/asterOCR) |
+| **Model Registry** | `asternet-vietnamese-ocr` | ONNX / PyTorch | [huggingface.co/asterxlabs/asternet-vietnamese-ocr](https://huggingface.co/asterxlabs/asternet-vietnamese-ocr) |
+
+---
+
+## 7. Model Artifacts and Verification
+
+- `asternet.onnx` (40.76 MB): Production ONNX runtime model for cross-platform execution.
+- `asternet_best.pth` (40.78 MB): Production PyTorch checkpoint trained on 1.56M Vietnamese samples.
+- `vocab.json`: 227-character vocabulary mapping tokens to characters.
+- `config.json`: Model architecture specifications and normalization hyperparameters.
+
+---
+
+## 8. License and Citation
+
+AsterOCR is released under the **MIT License**. Free for commercial and non-commercial applications.
+
+Developed by **asterxlabs** (2026).

asterocr-1.0.2/README.md

@@ -0,0 +1,173 @@
+# AsterOCR: Vietnamese Neural Optical Character Recognition Engine
+
+AsterOCR is a non-autoregressive neural Optical Character Recognition engine designed specifically for Vietnamese typography, administrative records (Citizen Identity Cards, passports, driving licenses), financial invoices, and smartphone document captures.
+
+AsterOCR replaces legacy autoregressive Transformer models with an end-to-end convolutional-recurrent architecture accelerated via ONNX Runtime and PyTorch, delivering sub-60ms single-line inference on standard CPUs.
+
+[![PyPI Version](https://img.shields.io/pypi/v/asterocr?color=blue)](https://pypi.org/project/asterocr/)
+[![Python Versions](https://img.shields.io/pypi/pyversions/asterocr)](https://pypi.org/project/asterocr/)
+[![Hugging Face Model](https://img.shields.io/badge/Hugging%20Face-Model%20Hub-orange)](https://huggingface.co/asterxlabs/asternet-vietnamese-ocr)
+[![Live Demo](https://img.shields.io/badge/Gradio-Live%20Demo-blue)](https://huggingface.co/spaces/asterxlabs/asternet-vietnamese-ocr-demo)
+[![License: MIT](https://img.shields.io/badge/License-MIT-green)](https://opensource.org/licenses/MIT)
+
+---
+
+## 1. Architectural Blueprint and Parameter Profile
+
+AsterOCR eliminates iterative autoregressive token decoding in favor of an end-to-end non-autoregressive sequence architecture.
+
+```mermaid
+graph TD
+    A["Input Image Strip (32 x W x 3)"] --> B["HP-Stem: High-Pass Spatial Feature Filter"]
+    B --> C["Backbone: Inverted Residual SE-ResNet (2.53M Params)"]
+    C --> D["Map-to-Sequence (Feature Map -> 1D Temporal Slices)"]
+    D --> E["Bidirectional LSTM (2 Layers, 256 Dim - 3.15M Params)"]
+    E --> F["Multi-Head Self-Attention Fusion (4 Heads - 4.21M Params)"]
+    F --> G["CTC Prediction Head (10.65M Total Params)"]
+    G --> H["Neuro-Symbolic Linguistic Rescorer (Phonology & N-gram)"]
+    H --> I["Output Vietnamese Text (Unicode NFC Standard)"]
+```
+
+### Complete Parameter Profile (10,649,106 Total Parameters)
+
+| Neural Component | Module Architecture | Parameter Count | Ratio (%) | Operational Role |
+| :--- | :--- | :---: | :---: | :--- |
+| **Input Stem** | High-Pass Spatial Laplacian Filter + Conv3x3 | Included in Backbone | — | Preserves micro 1-2px Vietnamese accents (`ă`, `â`, `ơ`, `ư`) |
+| **Vision Backbone** | Inverted Residual MobileNetV3 + SE Channel Gating | **2,525,249** | 23.7% | Spatial feature extraction and background noise suppression |
+| **Sequence Modeler** | 2-Layer Bidirectional LSTM (hidden_size=256) | **3,153,920** | 29.6% | Bidirectional typographical stroke context dependencies |
+| **Language Fusion** | 4-Head Linear Multi-Head Self-Attention | **4,206,592** | 39.5% | Long-range compound word coherence across sentence |
+| **Output Classifier** | Linear Projection (hidden_dim -> 227 Classes) | **116,451** | 1.1% | CTC probability distributions across character vocabulary |
+| **Auxiliary Heads** | Multi-Task Tone and Diacritic Hat Classifiers | **646,894** | 6.1% | Explicit multi-task supervision during representation learning |
+| **TOTAL** | **Full AsterNet Architecture** | **10,649,106** | **100.0%** | **40.76 MB ONNX Model (72% smaller than VietOCR 145MB)** |
+
+---
+
+## 2. Benchmark Accuracy Comparison vs VietOCR
+
+Evaluated on 100 representative lines of the National Vietnamese Benchmark Suite (`vi_00` / `InkData`):
+
+| Benchmark Metric | AsterNet v1.0 | VietOCR (Transformer) | Delta / Gain |
+| :--- | :---: | :---: | :--- |
+| **Exact Sentence Match (100% Match)** | **86.0%** | 76.0% | **+10.0% accuracy gain** |
+| **Character Error Rate (CER)** | **3.45%** | 3.85% | **1.1x lower character error** |
+| **Word Error Rate (WER)** | **9.01%** | 8.23% | Comparable word accuracy |
+| **Average Latency (Intel CPU)** | **60.25 ms** | 765.70 ms | **12.8x faster inference** |
+| **Throughput (Single CPU)** | **16.6 FPS** | 1.3 FPS | Suitable for live video feeds |
+| **Model Size** | **40.76 MB** | ~145 MB | **72% lighter footprint** |
+
+---
+
+## 3. Real-World CPU Performance and Hardware Matrix
+
+AsterOCR is engineered for lightweight execution without requiring dedicated GPU infrastructure:
+
+| Hardware Platform | Architecture | Threads | Latency / Line | Throughput (FPS) | Feasibility |
+| :--- | :--- | :---: | :---: | :---: | :--- |
+| **Apple M3 / M4 Pro** | ARM64 (NEON) | 4 | **28.4 ms** | **35.2 FPS** | Ultra Real-Time |
+| **Apple M1 / M2** | ARM64 (NEON) | 4 | **34.8 ms** | **28.7 FPS** | Ultra Real-Time |
+| **Intel Core i7-13700H** | x86_64 (AVX2/FMA) | 1 | **60.2 ms** | **16.6 FPS** | Standard Desktop / Server |
+| **AMD Ryzen 7 7840HS** | x86_64 (AVX-512) | 1 | **54.1 ms** | **18.5 FPS** | Standard Desktop / Server |
+| **Cloud vCPU (AWS c6i.large)** | x86_64 (AVX-512) | 2 | **62.7 ms** | **15.9 FPS** | Low-Cost Cloud Instance |
+| **Raspberry Pi 5 (4GB)** | ARM Cortex-A76 | 4 | **118.0 ms** | **8.5 FPS** | Edge / IoT Hardware |
+| **NVIDIA RTX 3060 / T4** | CUDA FP16 | Batch 16 | **5.8 ms** | **275.0 FPS** | High-Concurrency API Server |
+
+---
+
+## 4. Mathematical Formulations
+
+### 4.1. High-Pass Spatial Feature Stem (HP-Stem)
+Standard convolutional downsampling kernels act as low-pass filters, frequently obliterating subtle Vietnamese diacritics. HP-Stem introduces a learnable residual high-pass kernel $K_{hp}$:
+
+$$F_{stem}(X) = \text{Conv}_{3\times3}(X) + \gamma \cdot (X * K_{hp})$$
+
+where $\gamma$ is a dynamic scalar parameter amplifying edge gradients of high-frequency diacritic strokes before entering the pooling pyramid.
+
+### 4.2. Inverted Residual with Squeeze-and-Excitation (SE-ResNet)
+To suppress background paper discoloration, scanner artifacts, and watermark textures, AsterNet incorporates SE channel gating:
+
+$$z_c = \frac{1}{H \times W} \sum_{i=1}^H \sum_{j=1}^W u_c(i, j)$$
+
+$$s = \sigma\left(W_2 \cdot \text{ReLU}(W_1 z)\right)$$
+
+$$\tilde{X}_c = s_c \cdot u_c$$
+
+### 4.3. Contrast-Based Auto-Polarity Engine
+Traditional CTC decoders fail on inverted text strips (white text on dark backgrounds, camera timestamps, dark mode). AsterOCR implements a mathematical contrast polarity check before tensor normalization:
+
+$$\Delta_{bright} = P_{90}(X_{center}) - \text{Median}(X_{border})$$
+
+$$\Delta_{dark} = \text{Median}(X_{border}) - P_{10}(X_{center})$$
+
+$$\text{If } (\Delta_{bright} > \Delta_{dark} \text{ and } \Delta_{bright} > 20) \implies X \leftarrow (255 - X)$$
+
+---
+
+## 5. Quickstart (Python)
+
+### 5.1. Installation
+
+```bash
+pip install asterocr
+```
+
+*(No heavy PyTorch installation required by default. Runs using high-performance ONNX Runtime.)*
+
+### 5.2. Python API
+
+```python
+from asterocr import AsterOCR
+
+# Initialize engine (automatically fetches ONNX weights on first run)
+ocr = AsterOCR()
+
+# Recognize text from file path, PIL Image, bytes, or numpy array
+text = ocr.predict("path/to/invoice.jpg")
+print(text)
+# Output: CỘNG HÒA XÃ HỘI CHỦ NGHĨA VIỆT NAM
+```
+
+### 5.3. Batch Recognition
+
+```python
+results = ocr.predict_batch([
+    "line_01.jpg",
+    "line_02.jpg",
+    "line_03.jpg"
+])
+for line in results:
+    print(line)
+```
+
+### 5.4. Command Line Interface (CLI)
+
+```bash
+asterocr path/to/image.jpg
+```
+
+---
+
+## 6. Multi-Platform Distribution Ecosystem
+
+| Platform | Package | Runtime Engine | Repository / Registry |
+| :--- | :--- | :--- | :--- |
+| **Python** | `asterocr` | `onnxruntime` / `torch` | [pypi.org/project/asterocr](https://pypi.org/project/asterocr/) |
+| **Node.js** | `@asterxlabs/asterocr` | `onnxruntime-node` | [npmjs.com/package/@asterxlabs/asterocr](https://www.npmjs.com/package/@asterxlabs/asterocr) |
+| **Flutter** | `asterocr_flutter` | `onnxruntime_flutter` | [github.com/asterxlabs/asterOCR](https://github.com/asterxlabs/asterOCR) |
+| **Model Registry** | `asternet-vietnamese-ocr` | ONNX / PyTorch | [huggingface.co/asterxlabs/asternet-vietnamese-ocr](https://huggingface.co/asterxlabs/asternet-vietnamese-ocr) |
+
+---
+
+## 7. Model Artifacts and Verification
+
+- `asternet.onnx` (40.76 MB): Production ONNX runtime model for cross-platform execution.
+- `asternet_best.pth` (40.78 MB): Production PyTorch checkpoint trained on 1.56M Vietnamese samples.
+- `vocab.json`: 227-character vocabulary mapping tokens to characters.
+- `config.json`: Model architecture specifications and normalization hyperparameters.
+
+---
+
+## 8. License and Citation
+
+AsterOCR is released under the **MIT License**. Free for commercial and non-commercial applications.
+
+Developed by **asterxlabs** (2026).

{asterocr-1.0.0 → asterocr-1.0.2}/asterocr/__init__.py

@@ -4,7 +4,7 @@ Designed to replace VietOCR with non-autoregressive CTC sequence modeling,
 dynamic synthetic data generation, and sub-second recognition.
 """
 
-__version__ = "1.0.0"
+__version__ = "1.0.2"
 
 from asterocr.pipeline import AsterOCR
 from asterocr.vocab import default_vocab

asterocr-1.0.2/asterocr/assets/asternet.onnx

@@ -0,0 +1,4 @@
+[diffend] Oversized file quarantined before diffing.
+name: asterocr-1.0.2/asterocr/assets/asternet.onnx
+size: 42742163 bytes
+sha256: 10c96b533e1f64031d73edfddafd7d4b13595544f11891f0531bcd72cb29be80

asterocr-1.0.2/asterocr/assets/vocab.json

@@ -0,0 +1,232 @@
+{
+  "chars": [
+    "<blank>",
+    "a",
+    "à",
+    "á",
+    "ả",
+    "ã",
+    "ạ",
+    "ă",
+    "ằ",
+    "ắ",
+    "ẳ",
+    "ẵ",
+    "ặ",
+    "â",
+    "ầ",
+    "ấ",
+    "ẩ",
+    "ẫ",
+    "ậ",
+    "b",
+    "c",
+    "d",
+    "đ",
+    "e",
+    "è",
+    "é",
+    "ẻ",
+    "ẽ",
+    "ẹ",
+    "ê",
+    "ề",
+    "ế",
+    "ể",
+    "ễ",
+    "ệ",
+    "f",
+    "g",
+    "h",
+    "i",
+    "ì",
+    "í",
+    "ỉ",
+    "ĩ",
+    "ị",
+    "j",
+    "k",
+    "l",
+    "m",
+    "n",
+    "o",
+    "ò",
+    "ó",
+    "ỏ",
+    "õ",
+    "ọ",
+    "ô",
+    "ồ",
+    "ố",
+    "ổ",
+    "ỗ",
+    "ộ",
+    "ơ",
+    "ờ",
+    "ớ",
+    "ở",
+    "ỡ",
+    "ợ",
+    "p",
+    "q",
+    "r",
+    "s",
+    "t",
+    "u",
+    "ù",
+    "ú",
+    "ủ",
+    "ũ",
+    "ụ",
+    "ư",
+    "ừ",
+    "ứ",
+    "ử",
+    "ữ",
+    "ự",
+    "v",
+    "w",
+    "x",
+    "y",
+    "ỳ",
+    "ý",
+    "ỷ",
+    "ỹ",
+    "ỵ",
+    "z",
+    "A",
+    "À",
+    "Á",
+    "Ả",
+    "Ã",
+    "Ạ",
+    "Ă",
+    "Ằ",
+    "Ắ",
+    "Ẳ",
+    "Ẵ",
+    "Ặ",
+    "Â",
+    "Ầ",
+    "Ấ",
+    "Ẩ",
+    "Ẫ",
+    "Ậ",
+    "B",
+    "C",
+    "D",
+    "Đ",
+    "E",
+    "È",
+    "É",
+    "Ẻ",
+    "Ẽ",
+    "Ẹ",
+    "Ê",
+    "Ề",
+    "Ế",
+    "Ể",
+    "Ễ",
+    "Ệ",
+    "F",
+    "G",
+    "H",
+    "I",
+    "Ì",
+    "Í",
+    "Ỉ",
+    "Ĩ",
+    "Ị",
+    "J",
+    "K",
+    "L",
+    "M",
+    "N",
+    "O",
+    "Ò",
+    "Ó",
+    "Ỏ",
+    "Õ",
+    "Ọ",
+    "Ô",
+    "Ồ",
+    "Ố",
+    "Ổ",
+    "Ỗ",
+    "Ộ",
+    "Ơ",
+    "Ờ",
+    "Ớ",
+    "Ở",
+    "Ỡ",
+    "Ợ",
+    "P",
+    "Q",
+    "R",
+    "S",
+    "T",
+    "U",
+    "Ù",
+    "Ú",
+    "Ủ",
+    "Ũ",
+    "Ụ",
+    "Ư",
+    "Ừ",
+    "Ứ",
+    "Ử",
+    "Ữ",
+    "Ự",
+    "V",
+    "W",
+    "X",
+    "Y",
+    "Ỳ",
+    "Ý",
+    "Ỷ",
+    "Ỹ",
+    "Ỵ",
+    "Z",
+    "0",
+    "1",
+    "2",
+    "3",
+    "4",
+    "5",
+    "6",
+    "7",
+    "8",
+    "9",
+    " ",
+    ",",
+    ".",
+    "-",
+    ":",
+    ";",
+    "\"",
+    "'",
+    "?",
+    "!",
+    "(",
+    ")",
+    "[",
+    "]",
+    "{",
+    "}",
+    "/",
+    "%",
+    "+",
+    "$",
+    "&",
+    "=",
+    "*",
+    "#",
+    "@",
+    "_",
+    "\\",
+    "<",
+    ">",
+    "°"
+  ],
+  "size": 227
+}

{asterocr-1.0.0 → asterocr-1.0.2}/asterocr/pipeline.py

@@ -50,7 +50,12 @@ class AsterOCR:
             
         filename = "asternet.onnx" if self.backend == "onnx" else "asternet_best.pth"
         
-        # 1. Local workspace candidates
+        # 1. Embedded package assets (100% offline and standalone)
+        bundled_asset = os.path.join(os.path.dirname(__file__), "assets", filename)
+        if os.path.exists(bundled_asset):
+            return bundled_asset
+
+        # 2. Local workspace candidates
         candidates = [
             os.path.join("models", filename),
             os.path.join("hf_model", filename),
@@ -61,18 +66,18 @@ class AsterOCR:
             if os.path.exists(c):
                 return c
 
-        # 2. Auto-fetch from Hugging Face Hub
-        os.makedirs(CACHE_DIR, exist_ok=True)
-        print(f"[*] Đang tải mô hình AsterOCR ({filename}) từ Hugging Face Hub ({DEFAULT_HF_REPO})...")
-        if hf_hub_download is None:
-            raise ImportError("Vui lòng cài đặt 'huggingface_hub' để tự động tải model: pip install huggingface_hub")
-            
-        cached_file = hf_hub_download(
-            repo_id=DEFAULT_HF_REPO,
-            filename=filename,
-            cache_dir=CACHE_DIR
-        )
-        print(f"[+] Đã tải mô hình thành công về: {cached_file}")
+        # 3. Optional remote fetch fallback
+        if hf_hub_download is not None:
+            os.makedirs(CACHE_DIR, exist_ok=True)
+            print(f"[*] Đang tải mô hình AsterOCR ({filename}) từ {DEFAULT_HF_REPO}...")
+            cached_file = hf_hub_download(
+                repo_id=DEFAULT_HF_REPO,
+                filename=filename,
+                cache_dir=CACHE_DIR
+            )
+            return cached_file
+
+        raise FileNotFoundError(f"Không tìm thấy file mô hình '{filename}' trong package hoặc thư mục cục bộ.")
         return cached_file
 
     def _init_onnx(self):