timminator/VideOCR v1.4.2

VideOCR v1.4.2

on GitHub

Highlights:

• Switch to time based processing and complete rework of pipeline
  --> 2x Speed increase for Step 1
  --> Preprocessing step "Mapping" for VFR videos no longer necessary
  --> Enables instantaneous seeking
• Implements an interactive crop box with resizing and dragging
• Lowered default ocr_image_max_width value to speed up high-res video processing
• Add support for subtitle alignment tags (@quentin954)
• Adds a GUI scaling feature to manually increase font and button size
• Adds a preview of the specified brightness threshold to the Video Tab
• Optimized seeking in GUI
• Installed disk footprint reduced by 250MB

Fixes:

• Some buttons were not correctly updated with the selected language
• Fixes wrongly applying an already deleted crop box
• Adds a check to not allow editing of already running jobs
• Adds validation when adding jobs to queue

Main Highlights of the v1.4.x release:

• Integrates the latest PaddleOCR 3.4 version which extends support to 110 languages and provides increased accurary for latin, cyrillic, arabic etc.

Notes:

• Utilizes standalone version of PaddleOCR from https://github.com/timminator/PaddleOCR-Standalone
• Linux version requires glibc 2.35 or newer (can be checked by running ldd --version in a terminal)
• The --use_dual_zone parameter is removed starting from this version

Download Tips:

• Windows:
  • CPU version:
    VideOCR-CPU-v1.4.2-setup-x64.exe
  • GPU version:
    VideOCR-GPU-v1.4.2-CUDA-11.8-setup-x64.exe (Nvidia 10 Series graphics cards)
    VideOCR-GPU-v1.4.2-CUDA-12.9-setup-x64.exe (Nvidia 16 - 50 Series graphics cards)
• Linux:
  • CPU version:
    VideOCR-CPU-v1.4.2-Linux.7z
  • GPU version:
    VideOCR-GPU-v1.4.2-CUDA-11.8-Linux.7z (Nvidia 10 Series graphics cards)
    VideOCR-GPU-v1.4.2-CUDA-12.9-Linux.7z.001 (Nvidia 16 - 50 Series graphics cards)
    VideOCR-GPU-v1.4.2-CUDA-12.9-Linux.7z.002 (Nvidia 16 - 50 Series graphics cards)