Craniofacial Reconstruction Method Based on Region Fusion Strategy

Abstract

Craniofacial reconstruction is to estimate a person’s face model from the skull. It can be applied in many fields such as forensic medicine, archaeology, and face animation. Craniofacial reconstruction is based on the relationship between the skull and the face to reconstruct the facial appearance from the skull. However, the craniofacial structure is very complex and the relationship is not the same in different craniofacial regions. To better represent the shape changes of the skull and face and make better use of the correlation between different local regions, a new craniofacial reconstruction method based on region fusion strategy is proposed in this paper. This method has the flexibility of finding the nonlinear relationship between skull and face variables and is easy to solve. Firstly, the skull and face are divided into five corresponding local regions; secondly, the five regions of skull and face are mapped to low-dimensional latent space using Gaussian process latent variable model (GP-LVM), and the nonlinear features between skull and face are extracted; then, least square support vector regression (LSSVR) model is trained in latent space to establish the mapping relationship between skull region and face region; finally, perform regional fusion to achieve overall reconstruction. For the unknown skull, first divide the region, then project it into the latent space of the skull region, then use the trained LSSVR model to reconstruct the face of the corresponding region, and finally perform regional fusion to realize the face reconstruction of the unknown skull. The experimental results show that the method is effective. Compared with other regression methods, our method is optimal. In addition, we add attributes such as age and body mass index (BMI) to the mappings to achieve face reconstruction with different attributes.

Clinical applications of machine learning in predicting 3D shapes of the human body: a systematic review

“…Regarding the scope of this review, suitable methods applied in non-medical fields such as gaming and computer vision may have also been excluded. For example, in this systematic review studies in forensics were excluded since they were deemed to be out of the scope of this review [ 27 – 29 ]. Moreover, QUIPS checklist is not specifically developed for ML applications, however, it has been employed by other studies related to ML prediction models for clinical applications [ 30 , 31 ].…”

Contrastive Characteristic Analysis of Mandibular First Molar Alveolar Bone Between Li Nationality and Han Nationality

“…We speculate that Li nationality's mandible had its own characteristics in measurable and nonmeasurable traits before Qin Shihuang immigrated to Hainan Island in 221 BC. Jaw bone is the most complex and exquisite part of human craniofacial bones [7]. It is helpful to understand human physique, physiological state, living environment and culture at that time by studying the jaw bone [8].…”

3D craniofacial reconstruction framework using elastic surface deformation based on automatic landmark positioning

“…Facial organs not represented by a dry skull, such as nose, ears, and eyes, were added to the HRBF surface based on additional criteria. A similar approach to augmenting facial organs was introduced by Wen [5]. He uses a region fusion strategy to reconstruct facial organs.…”

“…This work selected one facial model as the facial template from the facial pool instead of a statistical model. As Wen et al [5] described, the statistical model is obtained through a potential relationship mining process to the skull-face sample database, which is statistically sufficient to represent a particular population.…”

“…They focused on only one of the stages of craniofacial reconstruction. Gietzen [1], Romeiro [4], and Wen [5] proposed a facial reconstruction method based on manually annotated landmarks. On the other hand, Alvarez [6], Jacinto [7], Lee [8], and Lindner [9] perform automatic landmark detection with 2D or 2.5D approaches.…”