The artificial neural networks have been established as a major paradigm in machine intelligence research. Their ability to learn from samples or instructions, supervised or unsupervised, and from binary or discrete inputs has made them an unbearable tool in dealing with complex problems.
Fields & Methodologies
- Feed-forward Multi-layered Nets
- Deep Architectures (CNNs, Autoencoders)
- Self-Organizing Maps
- Opposition-based ANNs
- Generaive vs. discriminative ANNs
Applications
- Big Image Data
- Image Identification
- Image Captioning
- Image Retireval