CMKD: CNN/Transformer-Based Cross-Model Knowledge Distillation for Audio Classification.
Yuan GongSameer KhuranaAndrew RouditchenkoJames R. GlassPublished in: CoRR (2022)
Keyphrases
- prior knowledge
- computational model
- probability distribution
- mathematical model
- multimedia
- feature vectors
- probabilistic model
- classification accuracy
- conceptual framework
- formal model
- additional knowledge
- classification models
- classification algorithm
- statistical model
- domain experts
- signal processing
- expert systems
- pattern recognition