scholarly journals Optimizing Few-Shot Learning Based on Variational Autoencoders

Entropy ◽  
2021 ◽  
Vol 23 (11) ◽  
pp. 1390
Author(s):  
Ruoqi Wei ◽  
Ausif Mahmood
Keyword(s):  
Face Recognition ◽  
Data Augmentation ◽  
Training Data ◽  
Training Dataset ◽  
The Novel ◽  
Learning Tasks ◽  
The Face ◽  
In The Wild ◽  
Face Generation ◽  
Generative Approach

Despite the importance of few-shot learning, the lack of labeled training data in the real world makes it extremely challenging for existing machine learning methods because this limited dataset does not well represent the data variance. In this research, we suggest employing a generative approach using variational autoencoders (VAEs), which can be used specifically to optimize few-shot learning tasks by generating new samples with more intra-class variations on the Labeled Faces in the Wild (LFW) dataset. The purpose of our research is to increase the size of the training dataset using various methods to improve the accuracy and robustness of the few-shot face recognition. Specifically, we employ the VAE generator to increase the size of the training dataset, including the basic and the novel sets while utilizing transfer learning as the backend. Based on extensive experimental research, we analyze various data augmentation methods to observe how each method affects the accuracy of face recognition. The face generation method based on VAEs with perceptual loss can effectively improve the recognition accuracy rate to 96.47% using both the base and the novel sets.

scholarly journals Optimizing Few-Shot Learning based on Variational Autoencoders

2021 ◽  
Author(s):  
Ruoqi Wei ◽  
Ausif Mahmood
Keyword(s):  
Face Recognition ◽  
Data Augmentation ◽  
Training Data ◽  
The Novel ◽  
Limited Data ◽  
Data Set ◽  
Learning Tasks ◽  
The Face ◽  
Face Generation ◽  
Generative Approach

Despite the importance of few-shot learning, the lack of labeled training data in the real world, makes it extremely challenging for existing machine learning methods as this limited data set does not represent the data variance well. In this research, we suggest employing a generative approach using variational autoencoders (VAEs), which can be used specifically to optimize few-shot learning tasks by generating new samples with more intra-class variations. The purpose of our research is to increase the size of the training data set using various methods to improve the accuracy and robustness of the few-shot face recognition. Specifically, we employ the VAE generator to increase the size of the training data set, including the basic and the novel sets while utilizing transfer learning as the backend. Based on extensive experimental research, we analyze various data augmentation methods to observe how each method affects the accuracy of face recognition. We conclude that the face generation method we proposed can effectively improve the recognition accuracy rate to 96.47% using both the base and the novel sets.

scholarly journals 3D-Aided Deep Pose-Invariant Face Recognition

2018 ◽  
Author(s):  
Jian Zhao ◽  
Lin Xiong ◽  
Yu Cheng ◽  
Yi Cheng ◽  
Jianshu Li ◽  
...  
Keyword(s):  
Face Recognition ◽  
Real Data ◽  
Training Data ◽  
Generative Adversarial Network ◽  
High Data ◽  
Adversarial Network ◽  
The Arts ◽  
The Face ◽  
In The Wild ◽  
Data Efficiency

Learning from synthetic faces, though perhaps appealing for high data efficiency, may not bring satisfactory performance due to the distribution discrepancy of the synthetic and real face images. To mitigate this gap, we propose a 3D-Aided Deep Pose-Invariant Face Recognition Model (3D-PIM), which automatically recovers realistic frontal faces from arbitrary poses through a 3D face model in a novel way. Specifically, 3D-PIM incorporates a simulator with the aid of a 3D Morphable Model (3D MM) to obtain shape and appearance prior for accelerating face normalization learning, requiring less training data. It further leverages a global-local Generative Adversarial Network (GAN) with multiple critical improvements as a refiner to enhance the realism of both global structures and local details of the face simulator’s output using unlabelled real data only, while preserving the identity information. Qualitative and quantitative experiments on both controlled and in-the-wild benchmarks clearly demonstrate superiority of the proposed model over state-of-the-arts.

A TIED-MIXTURE 2D HMM FACIAL IMAGE RETRIEVAL SYSTEM

2004 ◽  
Vol 13 (05) ◽  
pp. 1133-1146
Author(s):  
H. OTHMAN ◽  
T. ABOULNASR
Keyword(s):  
Face Recognition ◽  
Recognition System ◽  
Training Data ◽  
Expected Improvement ◽  
Model Case ◽  
Face Recognition System ◽  
Image Retrieval System ◽  
The Face ◽  
Institute Of Technology ◽  
Observation Space

In this paper, the effect of mixture tying on a second-order 2D Hidden Markov Model (HMM) is studied as applied to the face recognition problem. While tying HMM parameters is a well-known solution in the case of insufficient training data that leads to nonrobust estimation, it is used here to improve the overall performance in the small model case where the resolution in the observation space is the main problem. The fully-tied-mixture 2D HMM-based face recognition system is applied to the facial database of AT&T and the facial database of Georgia Institute of Technology. The performance of the proposed 2D HMM tied-mixture system is studied and the expected improvement is confirmed.

scholarly journals Improving Image-Based Plant Disease Classification With Generative Adversarial Network Under Limited Training Set

2020 ◽  
Vol 11 ◽  
Author(s):  
Luning Bi ◽  
Guiping Hu
Keyword(s):  
Classification Accuracy ◽  
Plant Disease ◽  
Data Augmentation ◽  
Plant Diseases ◽  
Disease Classification ◽  
Training Data ◽  
Training Dataset ◽  
Generative Adversarial Network ◽  
Adversarial Network ◽  
Overfitting Problem

Traditionally, plant disease recognition has mainly been done visually by human. It is often biased, time-consuming, and laborious. Machine learning methods based on plant leave images have been proposed to improve the disease recognition process. Convolutional neural networks (CNNs) have been adopted and proven to be very effective. Despite the good classification accuracy achieved by CNNs, the issue of limited training data remains. In most cases, the training dataset is often small due to significant effort in data collection and annotation. In this case, CNN methods tend to have the overfitting problem. In this paper, Wasserstein generative adversarial network with gradient penalty (WGAN-GP) is combined with label smoothing regularization (LSR) to improve the prediction accuracy and address the overfitting problem under limited training data. Experiments show that the proposed WGAN-GP enhanced classification method can improve the overall classification accuracy of plant diseases by 24.4% as compared to 20.2% using classic data augmentation and 22% using synthetic samples without LSR.

scholarly journals Clouds Classification from Sentinel-2 Imagery with Deep Residual Learning and Semantic Image Segmentation

2019 ◽  
Vol 11 (2) ◽  
pp. 119 ◽  
Author(s):  
Cheng-Chien Liu ◽  
Yu-Cheng Zhang ◽  
Pei-Yin Chen ◽  
Chien-Chih Lai ◽  
Yi-Hsin Chen ◽  
...  
Keyword(s):  
Image Segmentation ◽  
Change Detection ◽  
Data Augmentation ◽  
Satellite Image ◽  
Training Data ◽  
Training Dataset ◽  
Standard Size ◽  
Semantic Image Segmentation ◽  
Residual Learning ◽  
Sentinel 2

Detecting changes in land use and land cover (LULC) from space has long been the main goal of satellite remote sensing (RS), yet the existing and available algorithms for cloud classification are not reliable enough to attain this goal in an automated fashion. Clouds are very strong optical signals that dominate the results of change detection if they are not removed completely from imagery. As various architectures of deep learning (DL) have been proposed and advanced quickly, their potential in perceptual tasks has been widely accepted and successfully applied to many fields. A comprehensive survey of DL in RS has been reviewed, and the RS community has been suggested to be leading researchers in DL. Based on deep residual learning, semantic image segmentation, and the concept of atrous convolution, we propose a new DL architecture, named CloudNet, with an enhanced capability of feature extraction for classifying cloud and haze from Sentinel-2 imagery, with the intention of supporting automatic change detection in LULC. To ensure the quality of the training dataset, scene classification maps of Taiwan processed by Sen2cor were visually examined and edited, resulting in a total of 12,769 sub-images with a standard size of 224 × 224 pixels, cut from the Sen2cor-corrected images and compiled in a trainset. The data augmentation technique enabled CloudNet to have stable cirrus identification capability without extensive training data. Compared to the traditional method and other DL methods, CloudNet had higher accuracy in cloud and haze classification, as well as better performance in cirrus cloud recognition. CloudNet will be incorporated into the Open Access Satellite Image Service to facilitate change detection by using Sentinel-2 imagery on a regular and automatic basis.

scholarly journals Face Verification with Multi-Task and Multi-Scale Features Fusion

2017 ◽  
Author(s):  
Xiaojun Lu ◽  
Yue Yang ◽  
Weilin Zhang ◽  
Qi Wang ◽  
Yang Wang
Keyword(s):  
Training Dataset ◽  
Face Verification ◽  
Deep Convolutional Neural Networks ◽  
Computation Cost ◽  
Face Representation ◽  
Features Fusion ◽  
Multi Scale ◽  
The Face ◽  
In The Wild ◽  
Deep Cnn

Face verification for unrestricted faces in the wild is a challenging task. This paper proposes a method based on two deep convolutional neural networks(CNN) for face verification. In this work, we explore to use identification signal to supervise one CNN and the combination of semi-verification and identification to train the other one. In order to estimate semi-verification loss at a low computation cost, a circle, which is composed of all faces, is used for selecting face pairs from pairwise samples. In the process of face normalization, we propose to use different landmarks of faces to solve the problems caused by poses. And the final face representation is formed by the concatenating feature of each deep CNN after PCA reduction. What's more, each feature is a combination of multi-scale representations through making use of auxiliary classifiers. For the final verification, we only adopt the face representation of one region and one resolution of a face jointing Joint Bayesian classifier. Experiments show that our method can extract effective face representation with a small training dataset and our algorithm achieves 99.71% verification accuracy on LFW dataset.

Makeup Invariant Face Recognition using Features from Accelerated Segment Test and Eigen Vectors

2017 ◽  
Vol 17 (01) ◽  
pp. 1750005 ◽  
Author(s):  
Aruna Bhat
Keyword(s):  
Face Recognition ◽  
Principal Component ◽  
Recognition System ◽  
Training Data ◽  
Data Sets ◽  
Security Breaches ◽  
Daunting Task ◽  
Eigen Values ◽  
The Face ◽  
Robust Face Recognition

A methodology for makeup invariant robust face recognition based on features from accelerated segment test and Eigen vectors is proposed. Makeup and cosmetic changes in face have been a major cause of security breaches since long time. It is not only difficult for human eyes to catch an imposter but also an equally daunting task for a face recognition system to correctly identify an individual owing to changes brought about in face due to makeup. As a crucial pre-processing step, the face is first divided into various segments centered on the eyes, nose, lips and cheeks. FAST algorithm is then applied over the face images. The features thus derived from the facial image act as the fiducial points for that face. Thereafter principal component analysis is applied over the set of fiducial points in each segment of every face image present in the data sets in order to compute the Eigen vectors and the Eigen values. The resultant principal component which is the Eigen vector with the highest Eigen value yields the direction of the features in that segment. The principal components thus obtained using fiducial points generated from FAST in each segment of the test and the training data are compared in order to get the best match or no match.

FACE RECOGNITION IN COLOUR IMAGES USING NEURAL NETWORKS AND GENETIC ALGORITHMS

2005 ◽  
Vol 05 (01) ◽  
pp. 55-67 ◽  
Author(s):  
STEPHEN KARUNGARU ◽  
MINORU FUKUMI ◽  
NORIO AKAMATSU
Keyword(s):  
Neural Networks ◽  
Face Recognition ◽  
Template Matching ◽  
Distance Measure ◽  
Facial Feature ◽  
Training Data ◽  
Segmentation Method ◽  
Facial Feature Extraction ◽  
Matching Process ◽  
The Face

In this paper, a system that can automatically detect and recognise frontal faces is proposed. Three methods are used for face recognition; neural network, template matching and distance measure. One of the main problems encountered when using neural networks for face recognition is insufficient training data. This problem arises because, in most cases, only one image per subject is available. Therefore, amongst the objectives is to solve this problem by "increasing" the data available from the original image using several preprocesses, for example, image mirroring, colour and edges information, etc. Moreover, template matching is not trivial because of differences in the template shapes and sizes. In this work, template matching is aided by a genetic algorithm to automatically test several positions around the target and automatically adjust the size of the template as the matching process progresses. Distance measure method depends heavily on good facial feature extraction results. The image segmentation method applied matches such demand. The face colour information is represented using YIQ and the XYZ colour spaces. The effectiveness of the proposed method is verified by performing computer simulations. Two sets of databases were used. Database1 consists of 267 faces from the Oulu university database and database2 (for comparision purposes) consists of 250 faces from the ORL database.

scholarly journals Transfer learning using AlexNet Convolutional Neural Network for Face Recognition

2020 ◽  
Vol 9 (11) ◽  
pp. 285-294
Keyword(s):  
Neural Network ◽  
Face Recognition ◽  
Transfer Learning ◽  
Data Augmentation ◽  
Recognition System ◽  
Training Data ◽  
Fine Tuning ◽  
Data Sets ◽  
Learning Method ◽  
Face Recognition System

This research is aimed to achieve high-precision accuracy and for face recognition system. Convolution Neural Network is one of the Deep Learning approaches and has demonstrated excellent performance in many fields, including image recognition of a large amount of training data (such as ImageNet). In fact, hardware limitations and insufficient training data-sets are the challenges of getting high performance. Therefore, in this work the Deep Transfer Learning method using AlexNet pre-trained CNN is proposed to improve the performance of the face-recognition system even for a smaller number of images. The transfer learning method is used to fine-tuning on the last layer of AlexNet CNN model for new classification tasks. The data augmentation (DA) technique also proposed to minimize the over-fitting problem during Deep transfer learning training and to improve accuracy. The results proved the improvement in over-fitting and in performance after using the data augmentation technique. All the experiments were tested on UTeMFD, GTFD, and CASIA-Face V5 small data-sets. As a result, the proposed system achieved a high accuracy as 100% on UTeMFD, 96.67% on GTFD, and 95.60% on CASIA-Face V5 in less than 0.05 seconds of recognition time.

scholarly journals An improved age invariant face recognition using data augmentation

2021 ◽  
Vol 10 (1) ◽  
pp. 179-191
Author(s):  
Kennedy Okokpujie ◽  
Samuel John ◽  
Charles Ndujiuba ◽  
Joke A. Badejo ◽  
Etinosa Noma- Osaghae
Keyword(s):  
Face Recognition ◽  
Network Architecture ◽  
Data Augmentation ◽  
Research Question ◽  
Absolute Error ◽  
Training Model ◽  
Face Features ◽  
The Face ◽  
Using Data ◽  
Geometric Changes

In spite of the significant advancement in face recognition expertise, accurately recognizing the face of the same individual across different ages still remains an open research question. Face aging causes intra-subject variations (such as geometric changes during childhood adolescence, wrinkles and saggy skin in old age) which negatively affects the accuracy of face recognition systems. Over the years, researchers have devised different techniques to improve the accuracy of age invariant face recognition (AIFR) systems. In this paper, the face and gesture recognition network (FG-NET) aging dataset was adopted to enable the benchmarking of experimental results. The FG-Net dataset was augmented by adding four different types of noises at the preprocessing phase in order to improve the trait aging face features extraction and the training model used at the classification stages, thus addressing the problem of few available training aging for face recognition dataset. The developed model was an adaptation of a pre-trained convolution neural network architecture (Inception-ResNet-v2) which is a very robust noise. The proposed model on testing achieved a 99.94% recognition accuracy, a mean square error of 0.0158 and a mean absolute error of 0.0637. The results obtained are significant improvements in comparison with related works.

Sign in / Sign up

Export Citation Format

Share Document