OPFaceNet: OPtimized Face Recognition Network for noise and occlusion affected face images using Hyperparameters tuned Convolutional Neural Network.
Gurukumar LokkuGangireddy Harinatha ReddyM. N. Giri PrasadPublished in: Appl. Soft Comput. (2022)
Keyphrases
- face images
- face recognition
- hyperparameters
- convolutional neural network
- illumination variations
- pose variations
- noise level
- face detection
- facial expressions
- human faces
- principal component analysis
- face databases
- face recognition systems
- feature vectors
- face recognition algorithms
- cross validation
- recognition rate
- model selection
- feature extraction
- low resolution
- human face recognition
- face verification
- support vector
- ar face database
- training set
- sparse representation based classification
- computer vision
- closed form
- high resolution
- noisy images
- sparse representation
- bayesian framework
- bayesian inference
- noise reduction
- facial images
- random sampling
- object detection
- denoising
- multiscale
- high quality
- expectation maximization
- signal to noise ratio
- em algorithm