Robust Multi-Modal Group Action Recognition in Meetings from Disturbed Videos with the Asynchronous Hidden Markov Model.
Marc A. Al-HamesClaus LenzStephan ReiterJoachim SchenkFrank WallhoffGerhard RigollPublished in: ICIP (2) (2007)
Keyphrases
- multi modal
- hidden markov models
- action recognition
- view invariant
- human actions
- action classification
- recognition of human actions
- recognizing human actions
- audio visual
- video search
- gesture recognition
- bag of words
- human activities
- activity recognition
- semantic concepts
- space time interest points
- speech recognition
- action recognition in videos
- computer vision
- high dimensional
- spatio temporal
- conditional random fields
- multi view
- human motion
- speech signal
- bag of visual words
- action primitives
- body parts
- latent variables
- video content