 myhloli
|
ddf5a8781a
fix(batch): refactor OCR detection integration and area ratio calculation
|
il y a 5 mois |
|
myhloli
|
7c1d7dff65
fix(batch): disable OCR detection batch processing by default
|
il y a 5 mois |
|
Xiaomeng Zhao
|
5495055140
Update batch_analyze.py
|
il y a 5 mois |
 speta
|
99d4c97afc
支持batch-ocr-det，速度约提升3倍（200页pdf在3090上）
|
il y a 5 mois |
|
myhloli
|
1e01ffcf78
fix(ocr): adjust area ratio threshold and update fitz document handling in image conversion
|
il y a 5 mois |
|
myhloli
|
a2b848136b
feat(ocr): add area ratio calculation for OCR results and enhance get_coords_and_area function
|
il y a 5 mois |
|
myhloli
|
1d1c7ba9ab
refactor(table): replace ocr_engine with lang in table model prediction
|
il y a 7 mois |
|
myhloli
|
d2fc9dabf4
refactor(model): optimize batch processing and inference
|
il y a 7 mois |
|
myhloli
|
f8323ae07c
refactor(ocr): comment out det_count update and update OCR models
|
il y a 7 mois |
|
myhloli
|
df7ae4042d
feat(model): improve table recognition by merging and filtering tables
|
il y a 7 mois |
|
myhloli
|
ea730ae2e9
refactor(ocr): improve OCR score precision to three decimal places
|
il y a 7 mois |
|
myhloli
|
1fd72f5f3a
refactor(magic_pdf): optimize table recognition and layout detection
|
il y a 7 mois |
|
myhloli
|
553f250fc7
refactor(magic_pdf): optimize code and improve logging
|
il y a 7 mois |
|
myhloli
|
8e1c23393c
feat(model): add tqdm progress bar to model prediction loops
|
il y a 7 mois |
|
myhloli
|
59d6b195b0
refactor(model): integrate AtomModelSingleton for OCR and improve OCR result handling
|
il y a 7 mois |
|
myhloli
|
d7d85a2881
feat(ocr): implement language-specific OCR processing
|
il y a 7 mois |
|
myhloli
|
a330651d64
feat(ocr): implement separate detection and recognition processes
|
il y a 7 mois |
 icecraft
|
bbba2a120c
feat: batch inference with ocr and lang flag
|
il y a 7 mois |
|
myhloli
|
1b34f7e4ff
refactor(magic_pdf): replace PIL with NumPy for image processing
|
il y a 8 mois |
|
myhloli
|
49d140c55b
perf(model): adjust batch size for layout and formula detection
|
il y a 10 mois |
|
myhloli
|
452a9c0bcd
refactor(model): update batch analyze logic for rapid table model
|
il y a 10 mois |
|
myhloli
|
f350222614
feat(model): improve batch analysis logic and support npu
|
il y a 10 mois |
|
myhloli
|
902dcd2cee
refactor(BatchAnalyze): comment out image rotation logic in doclayout_yolo
|
il y a 10 mois |
|
myhloli
|
50f4841716
refactor(device): optimize memory cleaning and device selection
|
il y a 10 mois |
|
icecraft
|
b2887ca0aa
refactor: refactor code
|
il y a 11 mois |
 Suven
|
be01039495
feat: add logging for detection time in BatchAnalyze when OCR is not applied
|
il y a 11 mois |
|
Suven
|
49bfdf07c8
feat: enhance batch processing in BatchAnalyze with layout and OCR timing logs
|
il y a 11 mois |
|
Suven
|
4fd1e41e0e
fix: batch methods in DocLayoutYOLO and YOLOv8 models
|
il y a 11 mois |
|
Suven
|
7ce9edc674
feat: add batch prediction methods for YOLOv8 and Unimernet models
|
il y a 11 mois |
