Extract tables from images (PNG, JPEG, WebP, …) as structured JSON, Markdown, HTML or CSV. AE3-TL combines a deep-learning table-structure model (ONNX Runtime) with Tesseract OCR to reconstruct the ...