Showing 1–2 of 2 results for author: Hoshen, D

MakeupBag: Disentangling Makeup Extraction and Application

Authors: Dokhyam Hoshen

Abstract: This paper introduces MakeupBag, a novel method for automatic makeup style transfer. Our proposed technique can transfer a new makeup style from a reference face image to another previously unseen facial photograph. We solve makeup disentanglement and facial makeup application as separable objectives, in contrast to other current deep methods that entangle the two tasks. MakeupBag presents a signi… ▽ More This paper introduces MakeupBag, a novel method for automatic makeup style transfer. Our proposed technique can transfer a new makeup style from a reference face image to another previously unseen facial photograph. We solve makeup disentanglement and facial makeup application as separable objectives, in contrast to other current deep methods that entangle the two tasks. MakeupBag presents a significant advantage for our approach as it allows customization and pixel specific modification of the extracted makeup style, which is not possible using current methods. Extensive experiments, both qualitative and numerical, are conducted demonstrating the high quality and accuracy of the images produced by our method. Furthermore, in contrast to most other current methods, MakeupBag tackles both classical and extreme and costume makeup transfer. In a comparative analysis, MakeupBag is shown to outperform current state-of-the-art approaches. △ Less

Submitted 3 December, 2020; originally announced December 2020.

IQ of Neural Networks

Authors: Dokhyam Hoshen, Michael Werman

Abstract: IQ tests are an accepted method for assessing human intelligence. The tests consist of several parts that must be solved under a time constraint. Of all the tested abilities, pattern recognition has been found to have the highest correlation with general intelligence. This is primarily because pattern recognition is the ability to find order in a noisy environment, a necessary skill for intelligen… ▽ More IQ tests are an accepted method for assessing human intelligence. The tests consist of several parts that must be solved under a time constraint. Of all the tested abilities, pattern recognition has been found to have the highest correlation with general intelligence. This is primarily because pattern recognition is the ability to find order in a noisy environment, a necessary skill for intelligent agents. In this paper, we propose a convolutional neural network (CNN) model for solving geometric pattern recognition problems. The CNN receives as input multiple ordered input images and outputs the next image according to the pattern. Our CNN is able to solve problems involving rotation, reflection, color, size and shape patterns and score within the top 5% of human performance. △ Less

Submitted 29 September, 2017; originally announced October 2017.