Acral melanoma is the most common type of melanoma in Asians, and usually results in a poor prognosis due to late diagnosis. We applied a convolutional neural network to dermoscopy images of acral melanoma and benign nevi on the hands and feet and evaluated its usefulness for the early diagnosis of these conditions. A total of 724 dermoscopy images comprising acral melanoma (350 images from 81 patients) and benign nevi (374 images from 194 patients), and confirmed by histopathological examination, were analyzed in this study. To perform the 2-fold cross validation, we split them into two mutually exclusive subsets: half of the total image dataset was selected for training and the rest for testing, and we calculated the accuracy of diagnosis comparing it with the dermatologist’s and non-expert’s evaluation. The accuracy (percentage of true positive and true negative from all images) of the convolutional neural network was 83.51% and 80.23%, which was higher than the non-expert’s evaluation (67.84%, 62.71%) and close to that of the expert (81.08%, 81.64%). Moreover, the convolutional neural network showed area-under-the-curve values like 0.8, 0.84 and Youden’s index like 0.6795, 0.6073, which were similar score with the expert. Although further data analysis is necessary to improve their accuracy, convolutional neural networks would be helpful to detect acral melanoma from dermoscopy images of the hands and feet.

Facial expression recognition is one of the most important tasks in human–computer interaction, affective computing, computer vision, and related work. Feature additive pooling and progressive fine‐tuning of the convolutional neural network (CNN) for facial expression recognition in a static image are introduced. Network is proposed that partially employs the visual geometry group (VGG)‐face model pre‐trained on a VGG‐face dataset. The characteristics and distribution of the facial expression images in each database are biased according to the purpose of the publicly available facial database used. To alleviate this problem, a CNN model is developed that merges progressively fine‐tuned CNNs into a single network. Experiments were carried out to validate the presented method using facial expression images from the Cohn–Kanade +, Karolinska directed emotional face, and Japanese female facial expression databases, and cross‐database evaluation results show that the method is superior to state‐of‐the‐art methods.

We present photometric stereo algorithms robust to non-Lambertian reflection, which are based on a convolutional neural network in which surface normals of objects with complex geometry and surface reflectance are estimated from a given set of an arbitrary number of images. These images are taken from the same viewpoint under different directional illumination conditions. The proposed method focuses on surface normal estimation, where multi-scale feature aggregation is proposed to obtain a more accurate surface normal, and max pooling is adopted to obtain an intermediate order-agnostic representation in the photometric stereo scenario. The proposed multi-scale feature aggregation scheme using feature concatenation is easily incorporated into existing photometric stereo network architectures. Our experiments were performed with a DiLiGent photometric stereo benchmark dataset consisting of ten real objects, and they demonstrated that the accuracies of our calibrated and uncalibrated photometric stereo approaches were improved over those of baseline methods. In particular, our experiments also demonstrated that our uncalibrated photometric stereo outperformed the state-of-the-art method. Our work is the first to consider the multi-scale feature aggregation in photometric stereo, and we showed that our proposed multi-scale fusion scheme estimated the surface normal accurately and was beneficial to improving performance.

We present a reliable and accurate global optimization framework for estimating parameters of isotropic analytical bidirectional reflectance distribution function (BRDF) models. This approach is based on a branch and bound strategy with linear programming and interval analysis. Conventional local optimization is often very inefficient for BRDF estimation since its fitting quality is highly dependent on initial guesses due to the nonlinearity of analytical BRDF models. The algorithm presented in this paper employs $L_1$-norm error minimization to estimate BRDF parameters in a globally optimal way and interval arithmetic to derive our feasibility problem and lower bounding function. Our method is developed for the Cook–Torrance model but with several normal distribution functions such as the Beckmann, Berry, and GGX functions. Experiments have been carried out to validate the presented method using 100 isotropic materials from the MERL BRDF database, and our experimental results demonstrate that the $L_1$-norm minimization provides a more accurate and reliable solution than the $L_2$-norm minimization.

We present a sequential fitting-and-separating algorithm for surface reflectance components that separates individual dominant reflectance components and simultaneously estimates the corresponding bidirectional reflectance distribution function (BRDF) parameters from the separated reflectance values. We tackle the estimation of a Lafortune BRDF model, which combines a nonLambertian diffuse reflection and multiple specular reflectance components with a different specular lobe. Our proposed method infers the appropriate number of BRDF lobes and their parameters by separating and estimating each of the reflectance components using an interval analysis-based branch-and-bound method in conjunction with iterative K-ordered scale estimation. The focus of this paper is the estimation of the Lafortune BRDF model. Nevertheless, our proposed method can be applied to other analytical BRDF models such as the Cook–Torrance and Ward models. Experiments were carried out to validate the proposed method using isotropic materials from the Mitsubishi Electric Research Laboratories–Massachusetts Institute of Technology (MERL–MIT) BRDF database, and the results show that our method is superior to a conventional minimization algorithm.

A robust fitting algorithm for geometric computer vision problems under the L _∞ ‐norm optimisation framework is presented. It is essentially based on the maximum feasible subsystem (MaxFS) but it overcomes the computational limitation of the MaxFS for large data by finding only a quasi‐maximum feasible subset. Experimental results demonstrate that the algorithm removes outliers more effectively than the other parameter estimation methods recently developed when the outlier‐to‐inlier ratio in a data set is high.