Skip to content

CHANGELOG.md

Latest commit

 

History

History
138 lines (97 loc) · 4.16 KB

CHANGELOG.md

File metadata and controls

138 lines (97 loc) · 4.16 KB

Changelog

All notable changes to the CryptoNets Python SDK will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[2.0.3] - 2026-04-06

Changed

  • Updated native SDK binaries to version 26.04.03-833ba10
  • No API changes - fully backward compatible with 2.0.2

[2.0.2] - 2026-01-07

Fixed

  • Fixed use-after-free bug by copying buffer to bytes before freeing C-allocated memory as ffi.buffer() doest not copy the internal CData buffer
  • Fixed Native SDK version string initialization (was returned empty)

[2.0.0 (removed)-2.0.1] - 2026-01-05

Changed

Configuration API Updates

This release updates the SDK to align with the Generic V2 native library. Several configuration parameters have been modernized and enhanced:

Face Detection and Selection

  • BREAKING: Replaced single_face and consider_biggest_face boolean parameters with unified face_detection_strategy integer parameter
    • 0 = Multiple faces (all detected faces returned)
    • 1 = Best confidence score (single face with highest confidence) - default
    • 2 = Biggest face (single face with largest area)
    • 3 = Hybrid (best score of area × confidence)

Anti-Spoofing Configuration

  • BREAKING: Replaced skip_antispoof and use_jdb_antispoof boolean parameters with unified anti_spoofing_mode integer parameter
    • 0 = Off (no anti-spoofing)
    • 1 = XMS (dual XMS models, works with head pose landmarks) - default
    • 2 = JPD (JPD model, works with yolov5n_05_float16 landmarks)
    • 3 = Recognito Android (for Android platforms)

Face Landmarks Configuration

  • Generalized the face landmark selection for age to all operations to be configured via base_face_landmarks_model_id and age_face_landmarks_model_id was removed.
  • Added base_landmarks_model_id parameter for selecting face landmark detection model
    • 0 = Head pose model (default)
    • 22 = Yolov5n05 model

Added

  • landmark_confidence_score_threshold configuration parameter (default: 0.5) for face landmarks detection
  • base_landmarks_model_id configuration for selecting between head pose and Yolov5n05 models
  • Documentation for barcode detection features (previously marked as "future versions")

Documentation

  • Updated README.md to reflect Generic V2 native library capabilities
  • Clarified face landmark model configuration section
  • Removed "for future versions" notes for available features
  • Improved configuration parameter descriptions with model compatibility notes

Migration Guide

If you're upgrading from 2.0.0b1:

Face Selection Migration:

# Old (2.0.0b1)
config = OperationConfig(
    single_face=True,
    consider_biggest_face=False
)

# New (2.0.1)
config = OperationConfig(
    face_detection_strategy=1  # Best confidence score
)

# Equivalent mappings:
# single_face=True → face_detection_strategy=1 (best confidence)
# consider_biggest_face=True → face_detection_strategy=2 (biggest face)
# Multiple faces → face_detection_strategy=0

Anti-Spoofing Migration:

# Old (2.0.0b1)
config = OperationConfig(
    skip_antispoof=False,
    use_jdb_antispoof=False
)

# New (2.0.1)
config = OperationConfig(
    anti_spoofing_mode=1  # XMS mode (default)
)

# Equivalent mappings:
# skip_antispoof=True → anti_spoofing_mode=0 (off)
# use_jdb_antispoof=True → anti_spoofing_mode=2 (JPD)
# Default → anti_spoofing_mode=1 (XMS)

Face Landmarks Migration:

# Old (2.0.0b1)
config = OperationConfig(
    age_face_landmarks_model_id=-1
)

# New (2.0.1)
config = OperationConfig(
    base_face_landmarks_model_id=-1  # Use default model
    # OR explicitly set:
    # base_landmarks_model_id=0   # Head pose model
    # base_landmarks_model_id=22  # Yolov5n05 model
)

[2.0.0b1] - 2025-12-16

Added

  • Initial beta release of version 2.0.0
  • Support for Generic V2 native library features

Earlier Versions

For changes in versions prior to 2.0.0b1, please refer to V1 Documentation: SDK Docs