This client is intended for people with some familiarity with python. If you are unfamiliar with python try out our web application.
The SyncAiAnimationClient class in animation_client.py provides a client for the Emotech Sync AI Platform API to generate csv and fbx animation curve files. To generate content using the API a user needs their token, which can be found in the Your Token section in Account Settings on the website.
For a detailed overview of the arguments please see the API Documentation.
Create a virtual environment and install the required packages from requirements.txt.
pip install -r requirements.txt
Single requests can be made from the command line using single_request.py, while multiple requests can be made using multiple_requests_from_file.py providing a json file of the job specifications for these requests.
SSML tags can be used within the text field to control the prosody of the generated speech, and also the expression on the avatar. See the API Documentation for more details.
When using single_request.py the generated file will be saved to the output_file.
Some examples of usage of single_requests.py are shown below:
python single_request.py --language en-US --token USERS_TOKEN --output_file animation_curve.csv --output_type csv --text "A text for the sync AI platform" --actor female
python single_request.py --language en-US --token USERS_TOKEN --output_file animation_curve.fbx --output_type fbx --target_rig arkit --emotion happy --emotion_level 1 --text "A text for the sync AI platform" --actor male
Using an audio file that is on the local machine. The text transcript should match the audio in the file:
python single_request.py --language en-US --token USERS_TOKEN --output_file animation_curve.csv --output_type csv --audio_file /local/path/to/audio_file.wav --text "A text for the sync AI platform"
Using a url to an audio file hosted online. The text transcript should match the audio in the file:
python single_request.py --language en-US --token USERS_TOKEN --output_file animation_curve.csv --output_type csv --audio_url https://www.awebsite.com/audio_file.wav --text "A text for the sync AI platform"
Using ssml tags in the text:
python single_request.py --language en-US --token USERS_TOKEN --output_file animation_curve.csv --output_type csv --text "<speak>Here is a pause <break time='1s'/> and now <emo:express-as style='happy' styledegree='1.0'> the expression is happy</emo:express-as><\speak>" --actor female
inputs.json is an example json file specifying multiple jobs, which can be used with the client as shown in the example below.
python multiple_requests_from_file.py --token USERS_TOKEN --input_file inputs.json
In the above example, a json file of service responses will be saved to inputs_results.json.