Fully Integrated Analog Machine Learning Classifier Using Custom Activation Function for Low Resolution Image Classification.
Sanjeev Tannirkulam ChandrasekaranAkshay JayarajVinay Elkoori Ghantala KarnamImon BanerjeeArindam SanyalPublished in: IEEE Trans. Circuits Syst. I Regul. Pap. (2021)
Keyphrases
- low resolution
- fully integrated
- activation function
- image classification
- machine learning
- high resolution
- super resolution
- neural network
- decision trees
- support vector machine
- feature selection
- learning algorithm
- hidden layer
- low resolution images
- multi layer perceptron
- learning rate
- artificial neural networks
- feed forward
- pattern recognition
- back propagation
- feature extraction
- face images
- training data
- high quality
- neural nets
- multilayer perceptron
- image features
- image super resolution
- feature space
- machine learning algorithms
- signal processing
- basis functions
- text classification
- radial basis function
- computer vision
- class specific
- support vector
- network architecture
- supervised learning
- data analysis
- training phase
- sparse coding
- feature vectors
- training examples
- model selection
- support vector machine svm
- fuzzy neural network
- data mining
- data sets
- workflow management
- learning tasks