Visual NLP(Spark OCR) release notes 4.3.1

4.3.1

Release date: 17-02-2023

We’re glad to announce that Visual NLP 😎 4.3.1 has been released.

Highlights

• ImageTextCleaner & ImageTableDetector have improved memory consumption.
• New Annotators supported in LightPipelines.
• Table extraction from Digital PDFs pipeline now entirely supported as a LightPipeline.

ImageTextCleaner & ImageTableDetector improved memory consumption

• ImageTextCleaner & ImageTableDetector improved memory consumption: we reduced about 30% the memory consumption for this annotator making it more memory friendly and enabling running on memory constrained environments like Colab.

New Annotators supported in LightPipelines

Now the following annotators are supported in LightPipelines,

• PdfToHocr,
• HocrTokenizer,
• ImageTableDetector,
• ImageScaler,
• HocrToTextTable,

Table extraction from Digital PDFs pipeline now entirely supported as a LightPipeline.

• Our Table Extraction from digital PDFs pipeline now supports running as a LightPipeline, check the updated notebook: SparkOCRPdfToTable.ipynb

This release is compatible with Spark NLP for Healthcare 4.3.0, and Spark NLP 4.3.0.