Deep multiple instance learning for foreground speech localization in ambient audio from wearable devices.
Rajat HebbarPavlos PapadopoulosRamon ReyesAlexander F. DanversAngelina J. PolsinelliSuzanne A. MoseleyDavid A. SbarraMatthias R. MehlShrikanth NarayananPublished in: EURASIP J. Audio Speech Music. Process. (2021)
Keyphrases
- multiple instance learning
- wearable devices
- multiple instance
- multi class
- supervised learning
- multimedia
- image annotation
- semi supervised
- gaze tracking
- semi supervised learning
- background subtraction
- visual information
- moving objects
- image sequences
- class labels
- machine learning
- image retrieval
- gesture recognition
- object recognition
- feature selection