Version 0.10.0

Prebuilt deep-vision-processing lib version v0.10.0.

• Removed CUDA package builds from the GitHub Actions release workflow to avoid CI artifact and file-size limitations
• Dropped the FX2D renderer from all Processing examples and tests, simplifying setup and avoiding renderer-specific issues
• Bumped the library version to v0.10.0 and modernized the build to Java 17 (including CI)
• Updated the native and computer-vision dependency stack:
  • Gradle JavaCPP platform plugin: 1.5.8 -> 1.5.10
  • JavaCV: 1.5.8 -> 1.5.12
  • OpenCV: 4.6.0 -> 4.11.0
  • OpenBLAS: 0.3.21 -> 0.3.30
  • Tesseract: 5.2.0 -> 5.5.1
  • CUDA redist (when enabled): 11.8.x -> 12.9.x
• Switched Processing dependencies from bundled JARs to Maven artifacts:
  • Processing core 4.5.0
  • JOGL / Gluegen 2.5.0
  • Processing dependencies are now compileOnly (excluded from the fat jar, but available in tests)
• Updated MultiHumanPoseNetwork to match newer OpenCV APIs by explicitly specifying the algorithm hint for GaussianBlur