Hand-crafted versus learned representations for audio event detection.
Selver Ezgi KüçükbayAdnan YaziciSinan KalkanPublished in: Multim. Tools Appl. (2022)
Keyphrases
- event detection
- hand crafted
- deep learning
- domain independent
- linguistic features
- soccer video
- wordnet
- semantic role labeling
- knowledge engineering
- automatically generated
- activity recognition
- multimedia
- background knowledge
- supervised learning
- domain specific
- visual information
- unsupervised learning
- sports video
- video sequences
- machine learning
- part of speech
- multi modal
- image classification
- image retrieval
- feature selection