MLDT: Multi-task Learning with Denoising Transformer for Gait Identity and Emotion Recognition.
Weijie ShengXiaoyan LuXinde LiPublished in: AICCC (2021)
Keyphrases
- emotion recognition
- multi task learning
- denoising
- multi task
- audio visual
- gaussian processes
- learning tasks
- noisy images
- human computer interaction
- learning problems
- facial expressions
- high order
- theoretical analysis
- image processing
- sentiment analysis
- transfer learning
- facial images
- information fusion
- learning models
- multi class
- face recognition
- feature space
- data sets
- gaussian process
- emotional state
- learning algorithm
- training set
- multi modal
- decision trees
- text classification
- computer vision
- affective states