| |
|
|
| import torch |
| from torchvision import transforms |
| from PIL import Image |
|
|
| |
| amharic_chars = list(' ሀሁሂሃሄህሆለሉሊላሌልሎሐሑሒሓሔሕሖመሙሚማሜምሞሰሱሲሳስሶረሩሪራሬርሮሠሡሢሣሤሥሦሸሹሺሻሼሽሾቀቁቂቃቄቅቆበቡቢባቤብቦተቱቲታቴትቶቸቹቺቻቼችቾኀኃነኑኒናኔንኖኘኙኚኛኜኝኞአኡኢኣኤእኦከኩኪካኬክኮኸኹኺኻኼኽኾወዉዊዋዌውዎዐዑዒዓዔዕዖዘዙዚዛዜዝዞዠዡዢዣዤዥዦየዩዪያዬይዮደዱዲዳዴድዶጀጁጂጃጄጅጆገጉጊጋጌግጎጠጡጢጣጤጥጦጨጩጪጫጬጭጮጰጱጲጳጴጵጶጸጹጺጻጼጽጾፀፁፂፃፄፅፆፈፉፊፋፌፍፎፐፑፒፓፔፕፖቨቩቪቫቬቭቮ0123456789፥፣()-ሏሟሷሯሿቧቆቈቋቷቿኗኟዟዧዷጇጧጯጿፏኳኋኧቯጐጕጓ።') |
|
|
| char_to_idx = {char: idx + 1 for idx, char in enumerate(amharic_chars)} |
| char_to_idx['<UNK>'] = len(amharic_chars) + 1 |
| idx_to_char = {idx: char for char, idx in char_to_idx.items()} |
| idx_to_char[0] = '<blank>' |
|
|
| def preprocess_image(image: Image.Image) -> torch.Tensor: |
| """ |
| Preprocess the input image for the model. |
| """ |
| transform = transforms.Compose([ |
| transforms.Resize((224, 224)), |
| transforms.ToTensor(), |
| transforms.Normalize(mean=[0.485, 0.456, 0.406], |
| std=[0.229, 0.224, 0.225]), |
| ]) |
| return transform(image) |
|
|
| def decode_predictions(preds: torch.Tensor) -> str: |
| """ |
| Decode the model's predictions using Best Path Decoding. |
| """ |
| preds = torch.argmax(preds, dim=2).transpose(0, 1) |
| decoded_texts = [] |
| |
| for pred in preds: |
| pred = pred.cpu().numpy() |
| decoded = [] |
| previous = 0 |
| for p in pred: |
| if p != previous and p != 0: |
| decoded.append(idx_to_char.get(p, '<UNK>')) |
| previous = p |
| recognized_text = ''.join(decoded) |
| decoded_texts.append(recognized_text) |
| |
| return decoded_texts |
|
|
