AN ABSTRACT OF THE THESIS OF Xu Xu for the degree of Master of Science in Computer Science presented on June 13, 2017. Title: Classifying and Synthesizing 3D Shapes of Objects using Deep Neural Networks Abstract approved:
نویسندگان
چکیده
approved: Sinisa Todorovic Reasoning about 3D shape of objects is important for successful computer vision applications in robotics, 3D rendering and modeling. In this thesis, we address two problems – First, given an image, we generate 3D shape of the foreground object that appears in the image. Second, we predict the class label of the input 3D object shape. Recent work uses convolutional neural networks (CNNs) for these problems. However, their training is difficult, since it requires a large amount of training data. Also, in 3D shape generation, existing approaches can not generate realistic 3D shapes and have limited variations in generated 3D shapes. This thesis addresses these issues. We present two novel approaches, one for each problem. First, for 3D classification, we formulate CNN learning as a beam search aimed at identifying an optimal CNN architecture as well as estimating parameters of such an optimal CNN. This model pursuit approach is evaluated on 3D ShapeNet dataset. Second, we introduce a 3D-VAE-GAN (3D Variational AutoEncoder Generative Adversarial Network) model to synthesize high-quality 3D objects from 2D images in ObjectNet3D dataset. Our experimental evaluation shows that our new CNN learning achieves the state-of-the-art results with better modeling efficiency, i.e., with fewer parameters which are much easier to train. Also, in 3D shape synthesis, we achieve larger variability in shapes. c ©Copyright by Xu Xu June 13, 2017 All Rights Reserved Classifying and Synthesizing 3D Shapes of Objects using Deep Neural Networks
منابع مشابه
AN ABSTRACT OF THE THESIS OF Hani Alesaimi for the degree of Master of Science in Electrical and Computer Engineering presented on December 4, 2013. Title: Energy-Efcient Routing for Delay-Constrained Data Trafc in Linear \ireless Sensor Networks Abstract approved:
approved: Bechir Hamdaoui Linear wireless sensor networks (L\SN) are special class of wireless sensor networks where sensor nodes are deployed in a straight line. Monitoring industrial pipelines, railroads, tunnels, power lines, and borders are applications of L\SNs. \ireless sensors are tiny devices with limited energy resources; therefore, efcient energy routing in L\SNs is critical. In this ...
متن کاملLEARNING VIEWER-CENTERED PROJECTIONS FOR 3D SHAPE COMPLETION BY DAEYUN SHIN THESIS Submitted in partial fulfillment of the requirements for the degree of Master of Science in Computer Science in the Graduate College of the University of Illinois at Urbana-Champaign, 2017
The goal of this study is to determine the effectiveness of different 3D shape representations in learning to generate volumetric shapes using deep neural networks. We propose to automatically reconstruct a 3D model from a single-view image of an object by synthesizing multiple depth images and inferring the volume through multi-view 3D reconstruction. The final output is a 3D mesh inferred wit...
متن کاملHand Gesture Recognition from RGB-D Data using 2D and 3D Convolutional Neural Networks: a comparative study
Despite considerable enhances in recognizing hand gestures from still images, there are still many challenges in the classification of hand gestures in videos. The latter comes with more challenges, including higher computational complexity and arduous task of representing temporal features. Hand movement dynamics, represented by temporal features, have to be extracted by analyzing the total fr...
متن کاملImage Backlight Compensation Using Recurrent Functional Neural Fuzzy Networks Based on Modified Differential Evolution
In this study, an image backlight compensation method using adaptive luminance modification is proposed for efficiently obtaining clear images.The proposed method combines the fuzzy C-means clustering method, a recurrent functional neural fuzzy network (RFNFN), and a modified differential evolution.The proposed RFNFN is based on the two backlight factors that can accurately detect the compensat...
متن کاملJoint influence of leakage delays and proportional delays on almost periodic solutions for FCNNs
This paper deals with fuzzy cellular neural networks (FCNNs) with leakage delays and proportional delays. Applying the differential inequality strategy, fixed point theorem and almost periodic function principle, some sufficient criteria which ensure the existence and global attractivity of a unique almost periodic solution for fuzzy cellular neuralnetworks with leakage delays and p...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 2017