Skip to content

v1.2.0

Choose a tag to compare

@sparknv sparknv released this 17 Dec 05:27
· 4 commits to release/1.0 since this release

New Features

  • Add Network Preset option UI; provide a list of available services
    • List Mark, Claire, and James model service function ids by default
    • Customizable by editing nvcf_ace_presets.json
  • Add Service Status Indicator to the UI; a small circle with green, yellow, red, or grey
  • Add Import A2F Parameters to ACE menu; import json config to a AceAnimationPlayer node
  • Add Export A2F Service Parameters to ACE menu; export service configuration data from supported services
  • Add support of Audio2Face-3D Authoring service (Early Access users only)
    • Add client type option
    • Rename Request Animation button to Connect and Send Audio
    • Add interactive mode when connected to an authoring service
  • Add support of Maya playback caching
  • Add grpc error information to the system messages
  • Add emotion output and blendshape output to the output UI
  • Add input string clean up for using network attributes

Fixes and Changes

  • Change UI labels to support the Authoring Service
  • Change build system to cmake from premake
  • Remove python usage in the RequestSendAudio command
  • Fix an issue of out_of_breath emotion not affecting preferred emotion
  • Fix float precision inconsistency on the UI