Improved face recognition method using SVM-MRF with KTBD based KCM segmentation approach

Rangayya,; Virupakshappa,; Patil, Nagabhushan

doi:10.1007/s13198-021-01483-3

Cited by 13 publications

(3 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this paper, simulation data generated by smart communities and actual data generated by smart communities are respectively used to verify the performance of the algorithm in comparison with CNN-ALO-SVM, 5 SVM-MRF, 6 and DAG-SVM. 7 LibSVM 3.2.4 is used in the experiment as a simple, convenient, fast, and effective software package for SVM pattern recognition and regression.…”

Section: Experiments and Results Analysismentioning

confidence: 99%

An evolutionary intelligent data analysis in promoting smart community

Zhao

2023

Internet Technology Letters

View full text Add to dashboard Cite

Smart community construction is an integral part of smart city construction, and smart community management requires huge amounts of data as support. Currently, the data generated by some smart communities is scattered, and this data needs further analysis to realize value. This paper primarily studies data classification and parameter optimization. First, a novel K‐means clustering group support vector machines (SVM) method is proposed for data classification. For the parameter optimization problem of SVMs, evolutionary computation is used to seek the optimal solution through iterative evolution in a population composed of some feasible solutions. Then, the improved gray wolf optimization (iGWO) algorithm is used to optimize parameters and select features of SVM. Finally, to alleviate the situation that the minority samples are easily misjudged as noise samples due to the redundant features in the initial data, an oversampling method based on the iGWO and synthetic minority oversampling technique (SMOTE) is proposed, called iGWO–SMOTE–SVM. The experimental results demonstrate that the suggested approach on the six UCI datasets has acceptable accuracy, F1, and G‐Mean, which can well serve the construction of smart communities.

show abstract

Section: Experiments and Results Analysismentioning

confidence: 99%

An evolutionary intelligent data analysis in promoting smart community

Zhao

2023

Internet Technology Letters

View full text Add to dashboard Cite

show abstract

“…The AR-DB [63] comprises 2600 colour images in Bitmap Image File (.bmp) format of 100 people. Each person/subject is represented by 26 images with different facial expressions or configurations, including winking, centre-light, with/without spectacles, happy, sad, tired, normal, shocked, right-light, left-light and winking faces.…”

Section: Augmented Reality Face Database (Ar-db)mentioning

confidence: 99%

“…These include high-quality facial photos, controlled environmental settings, and practical algorithms that can adapt to variations in lighting, position, and emotional expressions. For research and evaluation purposes, the following face datasets are used: Augmented Reality [63], Extended Cohn-Kanade [64], Extended Yale B [65], and Enhanced Extended Yale B. These datasets offer diverse images for the development and testing of facial recognition algorithms.…”

Section: Datasets Descriptionmentioning

confidence: 99%

A Recurrent Feature Iterative Network For Suspect Identification using Convolutional and Long Short Term Memory Approach ⋆

Shree,

Mohapatra,

Suwanwiwat

et al. 2024

Preprint

View full text Add to dashboard Cite

In suspect identification systems, facial features play a crucial role in recognising individuals. However, the challenge lies in sustaining the accuracy of the system over a long period of time, ensuring that it remains consistently high, reliable, and effective. This research introduces a novel lightweight model that requires low trainable parameters, a significantly smaller number than pre-trained models, which use millions of trainable parameters. The newly proposed recurrent feature iterative network integrates a convolutional neural network and long short-term memory in a single structure to synthesise diverse images and to effectively extract facial features. The long short-term memory-based feature-recurrent system demonstrates a significant improvement in accuracy when tested on the augmented reality face database, enhanced extended Yale B, Cohn-Kanade and extended Yale B datasets, achieving encouraging accuracy rates of 99.23%, 99.16%, 98.99%, and 97.68%, respectively. These accuracies, in general, outperform traditional baselines of 68.65%. This research advances the field by providing an innovative solution to enhance suspect identification systems through advanced facial image feature extraction, resulting in significantly improved accuracy rates.

show abstract