v1.1.2

• Fixes for iOS 11:
  • Check if rendering to floating-point textures is available, and fall back to CPU if not
  • Added playsinline parameter to video elements, to get video in examples to play inline
• Added handling of dynamic resizing of video/camera input in examples
• Improved facedetection settings for better accuracy
• Added version to clm object

v1.1.1

• Updated examples:
  • use navigator.mediaDevices instead of navigator.getUserMedia to support Safari 11
  • handle different camera aspect ratios (with fix in #121 from @jitinl)
  • switch videos from .ogv to .webm format
  • cleanup
• PR #111 : proper resetting of positions (from @jpweeks)

v1.1.0

• Refactor code
• Add npm support
• Multiple fixes from PR #85
• Add common.js / node.js support to models, PR #102
• Switch to rollup for building
• Add es6 module build
• Add model_pca_20_svm.js model as default
• Reuse single web worker for face detection

v1.0.2

• Added web workers for face detection (PR #83). Enabled by default, but can be disabled via parameters.
• Fixed variable reference when running without mosse models (#77)
• Cleaned up redundant code + formatting

v1.0.1

• Fix for transparent PNGs (PR #45)
• Added event just before tracking starts (PR #53)
• Fixed typo in readme (PR #88)
• Use smaller canvas for face detection to avoid wrong detections + other fixes (PR #92)
• Moved examples and docs from gh-pages branch into main branches

Initial release

Initial release