Thai Trocr Thaigov V2

🚀 Thai Trocr Thaigov V2

This is a Vision Encoder Decoder Model. It can be used for Thai OCR tasks and is fine - tuned on a specific Thai dataset.

🚀 Quick Start

This model is a Vision Encoder Decoder Model with the following characteristics:

• It uses microsoft/trocr-base-handwritten as the encoder.
• It uses airesearch/wangchanberta-base-att-spm-uncased as the decoder.
• It is fine - tuned on a 250k synthetic text images dataset using the ThaiGov V2 Corpus.
• It uses SynthTIGER to generate synthetic text images.
• It is very useful for fine - tuning any Thai OCR task.

💻 Usage Examples

Basic Usage

from PIL import Image
from transformers import TrOCRProcessor, VisionEncoderDecoderModel

processor = TrOCRProcessor.from_pretrained("kkatiz/thai-trocr-thaigov-v2")
model = VisionEncoderDecoderModel.from_pretrained("kkatiz/thai-trocr-thaigov-v2")

image = Image.open("... your image path").convert("RGB")
pixel_values = processor(image, return_tensors="pt").pixel_values
generated_ids = model.generate(pixel_values)

generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
print(generated_text)

```markdown
![Model Widget](https://i.ibb.co/QXZFSNx/test7.png)
Output Text: รมว.ธรรมนัส ลงพื้นที่