Human Face Reconstruction from a Single Input Image Based on a Coupled Statistical Model

Three-dimensional human face modeling is a very popular topic with many applications,such as facial animation, face recognition and model-based facial video communication. Therefore, how to model the facial geometry, texture intensity and illumination variation is important in computer vision and graphics. Previous works on 3D head modeling from a single face image utilized prior information on 3D head models. However, it is difficult to accurately reconstruct the 3D face model from a single face image with expression since the facial expression induces 3D face model deformation in a complex manner. The main challenge is the coupling of the neutral 3D face model and the 3D deformation due to expression, thus making the 3D model estimation from a single face image with expression very challenging. On the other hand, the illumination condition also makes the problem more difficult. In this thesis,we focus on developing a 3D face model reconstruction system including surface registration and training of 3D face models with expressional deformations as well as the estimation of the 3D neutral shape and the 3D expressional deformation from a single face image. The proposed reconstruction algorithm integrates the linear and non-linear subspace representations for a prior 3D neutral morphable model and the probabilistic manifold-based 3D expressional deformation. We incorporate the face geometry, expression deformation, texture and illumination information into the problem so that it is well constrained. The reconstructed 3D face models can also be further extended and applied to many real-world applications.

In this paper, a novel framework for 3D face reconstruction from a single 2D face image was proposed. We focus on generating 3D face model without expensive devices and complicated calculation. First, we preprocess 2D face image, including illumination compensation, face detection and feature point extraction. Then, the method is based on a 3D morphable face model that encodes shape and texture in terms of model parameters. The prior 3D face model is a linear combination of “eigenheads” obtained by applying PCA to a training set of laser-scanned 3D faces. To account for pose and illumination variations, the algorithm simulates the process of image formation in 3D face and it estimates 3D face and texture of faces from a single image. As opposed to previous literature, our method can locate facial feature points automatically in a 2D facial image. Moreover, we also show that our proposed method for pose estimation can be successfully applied to 3D face reconstruction. In our experiment results, the method proposed has a satisfied performance regarding calculating time and the reconstructing photorealistic 3D faces.

This paper proposes a method for discovering the primary objects in single images by learning from videos in a purely unsupervised manner—the learning process is based on videos, but the generated network is able to discover objects from a single input image. The rough idea is that an image typically consists of multiple object instances (like the foreground and background) that have spatial transformations across video frames and they can be sparsely represented. By exploring the sparsity representation of a video with a neural network, one may learn the features of each object instance without any labels, which can be used to discover, recognize, or distinguish object instances from a single image. In this paper, we consider a relatively simple scenario, where each image roughly consists of a foreground and a background. Our proposed method is based on encoder-decoder structures to sparsely represent the foreground, background, and segmentation mask, which further reconstruct the original images. We apply the feed-forward network trained from videos for object discovery in single images, which is different from the previous co-segmentation methods that require videos or collections of images as the input for inference. The experimental results on various object segmentation benchmarks demonstrate that the proposed method extracts primary objects accurately and robustly, which suggests that unsupervised image learning tasks can benefit from the sparsity of images and the inter-frame structure of videos.

Most current three-dimensional reconstruction methods often require at least two or more input images to reconstruct the shape of the object. Traditionally, it is hard to reconstruct the 3D shape only from one single input image. In this paper, we propose an effective method to reconstruct the 3D texture surface from a single input image for the texture with similar appearances and reflectance properties. Exactly, the patch matching and optimization methods have been combined to resolve the under-constrained problem of 3D reconstruction from one single image. Experimental results have verified the effectiveness of the proposed method in 3D reconstruction according to realistic perception.

3D face modeling from 2D face images is of significant importance for face analysis, animation and recognition. Previous research on this topic mainly focused on 3D face modeling from a single 2D face image; however, a single face image can only provide a limited description of a 3D face. In many applications, for example, law enforcement, multi-view face images are usually captured for a subject during enrollment, which makes it desirable to build a 3D face texture model, given a pair of frontal and profile face images. We first determine the correspondence between un-calibrated frontal and profile face images through facial landmark alignment. An initial 3D face shape is then reconstructed from the frontal face image, followed by shape refinement utilizing the depth information provided by the profile image. Finally, face texture is extracted by mapping the frontal face image on the recovered 3D face shape. The proposed method is utilized for 2D face recognition in two scenarios: (i) normalization of probe image, and (ii) enhancing the representation capability of gallery set. Experimental results comparing the proposed method with a state-of-the-art commercial face matcher and densely sampled LBP on a subset of the FERET database show the effectiveness of the proposed 3D face texture model.

T3D face reconstruction from a single 2D image is mathematically ill-posed. However, to solve ill-posed problems in the area of computer vision, a variety of methods has been proposed; some of the solutions are to estimate latent information or to apply model based approaches. In this paper, we propose a novel method to reconstruct a 3D face from a single 2D face image based on pose estimation and a deformable model of 3D face shape. For 3D face reconstruction from a single 2D face image, it is the first task to estimate the depth lost by 2D projection of 3D faces. Applying the EM algorithm to facial landmarks in a 2D image, we propose a pose estimation algorithm to infer the pose parameters of rotation, scaling, and translation. After estimating the pose, much denser points are interpolated between the landmark points by a 3D deformable model and barycentric coordinates. As opposed to previous literature, our method can locate facial feature points automatically in a 2D facial image. Moreover, we also show that the proposed method for pose estimation can be successfully applied to 3D face reconstruction. Experiments demonstrate that our approach can produce reliable results for reconstructing photorealistic 3D faces.

We present a stochastic filtering approach to perform albedo estimation from a single non-frontal face image. Albedo estimation has far reaching applications in various computer vision tasks like illumination-insensitive matching, shape recovery, etc. We extend the formulation proposed in [3] that assumes face in known pose and present an algorithm that can perform albedo estimation from a single image even when pose information is inaccurate. 3D pose of the input face image is obtained as a byproduct of the algorithm. The proposed approach utilizes class-specific statistics of faces to iteratively improve albedo and pose estimates. Illustrations and experimental results are provided to show the effectiveness of the approach. We highlight the usefulness of the method for the task of matching faces across variations in pose and illumination. The facial pose estimates obtained are also compared against ground truth.

In this paper, we proposed an efficient framework for depth estimation from a single 2D face image to apply in 3D face reconstruction. The proposed framework consists of three components. Firstly, we recover the corresponding 2D images from 3D face models of database, and the feature points of these images will be extracted to form the 2D-3D database. Secondly, the same feature points extraction steps are executed to the test 2D image. Finally, according to the similarity of the triangles of the feature points between the test image and 2D-3D database, the depth of the test 2D face will be recovered. The experimental results show that the proposed method can estimate depth from a single 2D face image efficiently and robustly.

Portrait viewpoint and illumination editing is an important problem with several applications in VR/AR, movies, and photography. Comprehensive knowledge of geometry and illumination is critical for obtaining photorealistic results. Current methods are unable to explicitly model in 3D while handling both viewpoint and illumination editing from a single image. In this paper, we propose VoRF, a novel approach that can take even a single portrait image as input and relight human heads under novel illuminations that can be viewed from arbitrary viewpoints. VoRF represents a human head as a continuous volumetric field and learns a prior model of human heads using a coordinate-based MLP with individual latent spaces for identity and illumination. The prior model is learned in an auto-decoder manner over a diverse class of head shapes and appearances, allowing VoRF to generalize to novel test identities from a single input image. Additionally, VoRF has a reflectance MLP that uses the intermediate features of the prior model for rendering One-Light-at-A-Time (OLAT) images under novel views. We synthesize novel illuminations by combining these OLAT images with target environment maps. Qualitative and quantitative evaluations demonstrate the effectiveness of VoRF for relighting and novel view synthesis, even when applied to unseen subjects under uncontrolled illumination. This work is an extension of Rao et al. (VoRF: Volumetric Relightable Faces 2022). We provide extensive evaluation and ablative studies of our model and also provide an application, where any face can be relighted using textual input.

Image degraded by haze is a critical aspect in today's environment while getting a high-quality haze-free image remains an important task in computer vision. In recent year, many works have been done to improve the visibility of image taken under bad weather. Conventional designs use multiple image and/ or single image to deal with haze removal. In this paper, we use a simple but effective prior, a variation of distance (VoD) prior, to estimate the transmission map and remove haze from a single input image. The VoD prior is developed based on the idea that the outdoor visibility of images taken under hazy weather conditions seriously reduced when the distance increases. The thickness of the haze can be estimated effectively and a haze-free image can be recovered by adopting the VoD prior and the new haze imaging model. Our method is stable to image local regions containing objects in different depths. Our experiments showed that the proposed method achieved better results than several state-of-the-art methods, and it can be implemented very quickly. Our method due to its fast speed and the good visual effect is suitable for real-time applications. This work confirms that estimating the transmission map using the distance information instead the color information is a crucial point in image enhancement and especially single image haze removal.

We propose a novel framework to reconstruct super-resolution human shape from a single low-resolution input image. The approach overcomes limitations of existing approaches that reconstruct 3D human shape from a single image, which require high-resolution images together with auxiliary data such as surface normal or a parametric model to reconstruct high-detail shape. The proposed framework represents the reconstructed shape with a high-detail implicit function. Analogous to the objective of 2D image super-resolution, the approach learns the mapping from a low-resolution shape to its high-resolution counterpart and it is applied to reconstruct 3D shape detail from low-resolution images. The approach is trained end-to-end employing a novel loss function which estimates the information lost between a low and high-resolution representation of the same 3D surface shape. Evaluation for single image reconstruction of clothed people demonstrates that our method achieves high-detail surface reconstruction from low-resolution images without auxiliary data. Extensive experiments show that the proposed approach can estimate super-resolution human geometries with a significantly higher level of detail than that obtained with previous approaches when applied to low-resolution images. https://marcopesavento.github.io/SuRS/.

In this paper, we present a system for synthesizing 3D human face models containing different expressions from a single facial image. Given a frontal image of the target face with neutral expression, we first detect several key points denoting the shape of the face by Active Shape Model (ASM). Then we apply a RBF-based scattered data interpolation to reconstruct a 3D target face using a neutral expression 3D face model as reference. By analyzing a series of 3D expression face model, we segment the 3D reference model into regions automatically that each region is correspondent to a facial organ. From the expression set we construct a motion model for each facial action with respect to the target face in a local consistent manner. At last, the reconstructed 3D target face model with neutral expression and the facial action motion model are combined to generate 3D target face of various expressions. There are 3 contributions of our work: (1) We employ a set of registered 3D facial expression models as input, which enabled us to generate more complex and visual-realistic expressions than other parameter-based approaches and 2D image-based methods. (2) On the basis of a clustering-based segmentation, we developed a localized linear expression model, which make it possible for us to generate different facial expressions both locally and globally, thusly enlarge the space of synthesize output and break the limitation by the limited scale of the input expression model set. (3) A local space transform procedure is included that the output expression can fit distinct facial shapes regardless of the scarcity of variation of the facial shapes (fat or thin) in the input model set.

An effective method has been proposed in this paper to reconstruct the 3D shape from single texture image with similar appearances and reflectance properties. Two main steps are required: the first is the estimation of lighting parameters, which has been estimated by detecting the largest direction of the brightness change, the second is the 3D reconstruction from single input image, which has been reconstructed by combining the patch matching and optimization methods. Experiment results have verified the effectiveness of the proposed method according to realistic perception.

High dynamic range (HDR) imaging is an indispensable technique in modern photography. Traditional methods focus on HDR reconstruction from multiple images, solving the core problems of image alignment, fusion, and tone mapping, yet having a perfect solution due to ghosting and other visual artifacts in the reconstruction. Recent attempts at single-image HDR reconstruction show a promising alternative: by learning to map pixel values to their irradiance using a neural network, one can bypass the align-and-merge pipeline completely yet still obtain a high-quality HDR image. In this work, we propose a weakly supervised learning method that inverts the physical image formation process for HDR reconstruction via learning to generate multiple exposures from a single image. Our neural network can invert the camera response to reconstruct pixel irradiance before synthesizing multiple exposures and hallucinating details in under- and over-exposed regions from a single input image. To train the network, we propose a representation loss, a reconstruction loss, and a perceptual loss applied on pairs of under- and over-exposure images and thus do not require HDR images for training. Our experiments show that our proposed model can effectively reconstruct HDR images. Our qualitative and quantitative results show that our method achieves state-of-the-art performance on the DrTMO dataset. Our code is available at https://github.com/VinAIResearch/single_image_hdr.

Advances in high dynamic range (HDR) lighting estimation from a single image have opened new possibilities for augmented reality (AR) applications. Predicting complex lighting environments from a single input image allows for the realistic rendering and compositing of virtual objects. In this work, we investigate the color robustness of such methods—an often overlooked yet critical factor for achieving visual realism. While most evaluations conflate color with other lighting attributes (e.g., intensity, direction), we isolate color as the primary variable of interest. Rather than introducing a new lighting estimation algorithm, we explore whether simple adaptation techniques can enhance the color accuracy of existing models. Using a novel HDR dataset featuring diverse lighting colors, we systematically evaluate several adaptation strategies. Our results show that preprocessing the input image with a pre-trained white balance network improves color robustness, outperforming other strategies across all tested scenarios. Notably, this approach requires no retraining of the lighting estimation model. We further validate the generality of this finding by applying the technique to three state-of-the-art lighting estimation methods from recent literature. Our project webpage is available at: https: // lvsn. github. io/ coloraccuracy .