"Sheldon speaking, Bonjour!": Leveraging Multilingual Tracks for (Weakly) Supervised Speaker Identification.
Hervé BredinAnindya RoyNicolas PécheuxAlexandre AllauzenPublished in: ACM Multimedia (2014)
Keyphrases
- weakly supervised
- speaker identification
- speech recognition
- gaussian mixture model
- superpixels
- speech signal
- feature extraction
- noisy environments
- relation extraction
- object class
- broadcast news
- topic models
- semi supervised
- named entities
- object detectors
- object detection
- computer vision
- image processing
- em algorithm
- image segmentation
- pattern recognition