Multi-View Graph Fusion for Semi-Supervised Learning: Application to Image-Based Face Beauty Prediction

Abstract

Facial Beauty Prediction (FBP) is an important visual recognition problem to evaluate the attractiveness of faces according to human perception. Most existing FBP methods are based on supervised solutions using geometric or deep features. Semi-supervised learning for FBP is an almost unexplored research area. In this work, we propose a graph-based semi-supervised method in which multiple graphs are constructed to find the appropriate graph representation of the face images (with and without scores). The proposed method combines both geometric and deep feature-based graphs to produce a high-level representation of face images instead of using a single face descriptor and also improves the discriminative ability of graph-based score propagation methods. In addition to the data graph, our proposed approach fuses an additional graph adaptively built on the predicted beauty values. Experimental results on the SCUTFBP-5500 facial beauty dataset demonstrate the superiority of the proposed algorithm compared to other state-of-the-art methods.

Facial Beauty Standards Predictions Based on Machine Learning: A Comparative Analysis

“…Rather of relying just on a single face descriptor, the suggested approach incorporates both geometric and deep feature-based graphs to provide a high-level representation of face pictures. It also enhances the discriminative power of graph-based score propagation techniques is discussed in [23]. In [24] the approach involves creating a face identification framework that uses 2D facial photos collected from several sources to create a 3D face mesh with 468 Media Pipe markers that can identify numerous faces in real time.…”

Anchor-Net: Distance-Based Self-Supervised Learning Model for Facial Beauty Prediction

“…Semi-supervised learning models [25], [26], [29], [43], [44], including NFME [25] and MSMFME [26], underscore the value of graph-based and multi-view graph fusion techniques in reinforcing model training without additional labeled images. Despite their innovative approach, these models face limitations in computing similarity graphs before estimating beauty predictions, indicating a gap in capturing the relative aspects of beauty.…”

Facial Beauty Prediction Using an Ensemble of Deep Convolutional Neural Networks

“…The standard SCUT-FBP5500 dataset [15] is introduced in this study, which comprises 5,500 frontal face images at 350 × 350 resolutions with various attributes, including race (Asian/Caucasian), gender (female/male), and age (15-60) [23,24].…”