A preliminary study of challenges in extracting purity videos from the AV Speech Benchmark.
Haoran YanHuijun LuDunbo CaiTao HangLing QianPublished in: ICMIP (2022)
Keyphrases
- audio visual
- real world
- video sequences
- lessons learned
- key issues
- video clips
- visual data
- computer vision
- multi modal
- open issues
- dynamic scenes
- spoken language
- audio features
- emotion recognition
- speech signal
- neural network
- video surveillance
- dialogue system
- video content
- language acquisition
- news video
- video data
- pattern recognition
- case study
- broadcast news
- endpoint detection