scholarly journals Learning and transferring motion style using Sparse PCA

2019 ◽  
Vol 35 (1) ◽  
Author(s):  
Khac Phong Do ◽  
Nguyen Xuan Thanh ◽  
Hongchuan Yu
Keyword(s):  
Computer Animation ◽  
Data Driven ◽  
Large Dataset ◽  
Sparse Pca ◽  
Style Transfer ◽  
Novel Method ◽  
Motion Style ◽  
Primary Problem ◽  
Small Dataset

Motion style transfer is a primary problem in computer animation, allowing us to convert the motion of an actor to that of another one. Myriads approaches have been developed to perform this task, however, the majority of them are data-driven, which require a large dataset and a time-consuming period for training a model in order to achieve good results. In contrast, we propose a novel method applied successfully for this task in a small dataset. This exploits Sparse PCA to decompose original motions into smaller components which are learned with particular constraints. The synthesized results are highly precise and smooth motions with its emotion as shown in our experiments.

scholarly journals Learning Domain Invariant Representation via Self-Rugularization

2021 ◽  
Vol 24 (4) ◽  
pp. 382-391
Author(s):  
Jaeguk Hyun ◽  
ChanYong Lee ◽  
Hoseong Kim ◽  
Hyunjung Yoo ◽  
Eunjin Koh
Keyword(s):  
Domain Adaptation ◽  
Regularization Methods ◽  
Invariant Representation ◽  
Large Dataset ◽  
Style Transfer ◽  
Unsupervised Domain Adaptation ◽  
Loss Term ◽  
Target Data ◽  
Small Dataset ◽  
Feature Alignment

Unsupervised domain adaptation often gives impressive solutions to handle domain shift of data. Most of current approaches assume that unlabeled target data to train is abundant. This assumption is not always true in practices. To tackle this issue, we propose a general solution to solve the domain gap minimization problem without any target data. Our method consists of two regularization steps. The first step is a pixel regularization by arbitrary style transfer. Recently, some methods bring style transfer algorithms to domain adaptation and domain generalization process. They use style transfer algorithms to remove texture bias in source domain data. We also use style transfer algorithms for removing texture bias, but our method depends on neither domain adaptation nor domain generalization paradigm. The second regularization step is a feature regularization by feature alignment. Adding a feature alignment loss term to the model loss, the model learns domain invariant representation more efficiently. We evaluate our regularization methods from several experiments both on small dataset and large dataset. From the experiments, we show that our model can learn domain invariant representation as much as unsupervised domain adaptation methods.

Gaussian Process Emulation for Big Data in Data-Driven Metamaterials Design

2019 ◽  
Author(s):  
Ramin Bostanabad ◽  
Yu-Chin Chan ◽  
Liwei Wang ◽  
Ping Zhu ◽  
Wei Chen
Keyword(s):  
Big Data ◽  
Gaussian Process ◽  
Engineering Design ◽  
Training Data ◽  
Data Driven ◽  
Training Dataset ◽  
Massive Datasets ◽  
Unit Cells ◽  
Gaussian Process Emulation ◽  
Novel Method

Abstract Our main contribution is to introduce a novel method for Gaussian process (GP) modeling of massive datasets. The key idea is to build an ensemble of independent GPs that use the same hyperparameters but distribute the entire training dataset among themselves. This is motivated by our observation that estimates of the GP hyperparameters change negligibly as the size of the training data exceeds a certain level, which can be found in a systematic way. For inference, the predictions from all GPs in the ensemble are pooled to efficiently exploit the entire training dataset for prediction. We name our modeling approach globally approximate Gaussian process (GAGP), which, unlike most largescale supervised learners such as neural networks and trees, is easy to fit and can interpret the model behavior. These features make it particularly useful in engineering design with big data. We use analytical examples to demonstrate that GAGP achieves very high predictive power that matches or exceeds that of state-of-the-art machine learning methods. We illustrate the application of GAGP in engineering design with a problem on data-driven metamaterials design where it is used to link reduced-dimension geometrical descriptors of unit cells and their properties. Searching for new unit cell designs with desired properties is then accomplished by employing GAGP in inverse optimization.

scholarly journals A Novel Method for Early Gear Pitting Fault Diagnosis Using Stacked SAE and GBRBM

Sensors ◽  
2019 ◽  
Vol 19 (4) ◽  
pp. 758 ◽  
Author(s):  
Jialin Li ◽  
Xueyi Li ◽  
David He ◽  
Yongzhi Qu
Keyword(s):  
Fault Diagnosis ◽  
Time Domain ◽  
Continuous Time ◽  
Data Driven ◽  
Vibration Signals ◽  
Vibration Data ◽  
The Past ◽  
Diagnosis Accuracy ◽  
Novel Method ◽  
Faults Diagnosis

Research on data-driven fault diagnosis methods has received much attention in recent years. The deep belief network (DBN) is a commonly used deep learning method for fault diagnosis. In the past, when people used DBN to diagnose gear pitting faults, it was found that the diagnosis result was not good with continuous time domain vibration signals as direct inputs into DBN. Therefore, most researchers extracted features from time domain vibration signals as inputs into DBN. However, it is desirable to use raw vibration signals as direct inputs to achieve good fault diagnosis results. Therefore, this paper proposes a novel method by stacking spare autoencoder (SAE) and Gauss-Binary restricted Boltzmann machine (GBRBM) for early gear pitting faults diagnosis with raw vibration signals as direct inputs. The SAE layer is used to compress the raw vibration data and the GBRBM layer is used to effectively process continuous time domain vibration signals. Vibration signals of seven early gear pitting faults collected from a gear test rig are used to validate the proposed method. The validation results show that the proposed method maintains a good diagnosis performance under different working conditions and gives higher diagnosis accuracy compared to other traditional methods.

scholarly journals MOTION DEFORMATION STYLE CONTROL TECHNIQUE FOR 3D HUMANOID CHARACTER BY USING MOCAP DATA

2015 ◽  
Vol 78 (2-2) ◽  
Author(s):  
Ismahafezi Ismail ◽  
Mohd Shahrizal Sunar ◽  
Hoshang Kolivand
Keyword(s):  
Virtual Reality ◽  
Motion Capture ◽  
Computer Games ◽  
Computer Animation ◽  
Mixed Reality ◽  
Human Action ◽  
Control Technique ◽  
Interface Control ◽  
Motion Style ◽  
Animation System

Realistic humanoid 3D character movement is very important to apply in the computer games, movies, virtual reality and mixed reality environment. This paper presents a technique to deform motion style using Motion Capture (MoCap) data based on computer animation system. By using MoCap data, natural human action style could be deforming. However, the structure hierarchy of humanoid in MoCap Data is very complex. This method allows humanoid character to respond naturally based on user motion input. Unlike existing 3D humanoid character motion editor, our method produces realistic final result and simulates new dynamic humanoid motion style based on simple user interface control.

scholarly journals A Cyclic Consistency Motion Style Transfer Method Combined with Kinematic Constraints

2021 ◽  
Vol 2021 ◽  
pp. 1-17
Author(s):  
Huaijun Wang ◽  
Dandan Du ◽  
Junhuai Li ◽  
Wenchao Ji ◽  
Lei Yu
Keyword(s):  
Motion Capture ◽  
Human Movement ◽  
Motion Capture Data ◽  
Style Transfer ◽  
Kinematic Constraints ◽  
Film And Television ◽  
Adversarial Network ◽  
Transfer Method ◽  
Motion Style ◽  
The Cost

Motion capture technology plays an important role in the production field of film and television, animation, etc. In order to reduce the cost of data acquisition and improve the reuse rate of motion capture data and the effect of movement style migration, the synthesis technology of motion capture data in human movement has become a research hotspot in this field. In this paper, kinematic constraints (KC) and cyclic consistency (CC) network are employed to study the methods of kinematic style migration. Firstly, cycle-consistent adversarial network (CCycleGAN) is constructed, and the motion style migration network based on convolutional self-encoder is used as a generator to establish the cyclic consistent constraint between the generated motion and the content motion, so as to improve the action consistency between the generated motion and the content motion and eliminate the lag phenomenon of the generated motion. Then, kinematic constraints are introduced to normalize the movement generation, so as to solve the problems such as jitter and sliding step in the movement style migration results. Experimental results show that the generated motion of the cyclic consistent style transfer method with kinematic constraints is more similar to the style of style motion, which improves the effect of motion style transfer.

scholarly journals Spectral Shape Recovery and Analysis Via Data-driven Connections

2021 ◽  
Author(s):  
Riccardo Marin ◽  
Arianna Rampini ◽  
Umberto Castellani ◽  
Emanuele Rodolà ◽  
Maks Ovsjanikov ◽  
...  
Keyword(s):  
Shape Matching ◽  
Ad Hoc ◽  
Computational Cost ◽  
Point Clouds ◽  
Data Driven ◽  
Spectrum Estimation ◽  
Unified Framework ◽  
3D Vision ◽  
Style Transfer ◽  
Latent Space

AbstractWe introduce a novel learning-based method to recover shapes from their Laplacian spectra, based on establishing and exploring connections in a learned latent space. The core of our approach consists in a cycle-consistent module that maps between a learned latent space and sequences of eigenvalues. This module provides an efficient and effective link between the shape geometry, encoded in a latent vector, and its Laplacian spectrum. Our proposed data-driven approach replaces the need for ad-hoc regularizers required by prior methods, while providing more accurate results at a fraction of the computational cost. Moreover, these latent space connections enable novel applications for both analyzing and controlling the spectral properties of deformable shapes, especially in the context of a shape collection. Our learning model and the associated analysis apply without modifications across different dimensions (2D and 3D shapes alike), representations (meshes, contours and point clouds), nature of the latent space (generated by an auto-encoder or a parametric model), as well as across different shape classes, and admits arbitrary resolution of the input spectrum without affecting complexity. The increased flexibility allows us to address notoriously difficult tasks in 3D vision and geometry processing within a unified framework, including shape generation from spectrum, latent space exploration and analysis, mesh super-resolution, shape exploration, style transfer, spectrum estimation for point clouds, segmentation transfer and non-rigid shape matching.

scholarly journals Unpaired motion style transfer from video to animation

2020 ◽  
Vol 39 (4) ◽  
Author(s):  
Kfir Aberman ◽  
Yijia Weng ◽  
Dani Lischinski ◽  
Daniel Cohen-Or ◽  
Baoquan Chen
Keyword(s):  
Style Transfer ◽  
Motion Style

scholarly journals A Novel Method for Estimating Substitution Rate Variation Among Sites in a Large Dataset of Homologous DNA Sequences

Genetics ◽  
2001 ◽  
Vol 157 (2) ◽  
pp. 859-865
Author(s):  
Graziano Pesole ◽  
Cecilia Saccone
Keyword(s):  
Dna Sequences ◽  
Rate Variation ◽  
Large Dataset ◽  
Site Specific ◽  
Homologous Sequences ◽  
Large Sets ◽  
Hypervariable Regions ◽  
D Loop ◽  
Novel Method ◽  
Substitution Rate Variation

Abstract We present here a novel method to estimate the site-specific relative variability in large sets of homologous sequences. It is based on the simple idea that the more closely related are the compared sequences, the higher the probability of observing nucleotide changes at rapidly evolving sites. A simulation study has been carried out to support the reliability of the method, which has been applied also to analyzing the site variability of all available human sequences corresponding to the two hypervariable regions of the mitochondrial D-loop.

scholarly journals A data-driven method to predict achievability of clinical objectives in IMRT

2021 ◽  
Author(s):  
Vaitheeswaran Ranganathan
Keyword(s):  
Head And Neck ◽  
Treatment Planning ◽  
Region Of Interest ◽  
Target Volume ◽  
Data Driven ◽  
Planning System ◽  
Treatment Planning System ◽  
Geometric Complexity ◽  
Imrt Planning ◽  
Novel Method

Abstract When specifying a clinical objective for a target volume and normal organs/tissues in IMRT planning, the user may not be sure if the defined clinical objective could be achieved by the optimizer. To this end, we propose a novel method to predict the achievability of clinical objectives upfront before invoking the optimization. A new metric called “Geometric Complexity (GC)” is used to estimate the achievability of clinical objectives. Essentially GC is the measure of the number of “unmodulated” beamlets or rays that intersect the Region-of-interest (ROI) and the target volume. We first compute the geometric complexity ratio (GCratio) between the GC of a ROI in a reference plan and the GC of the same ROI in a given plan. The GCratio of a ROI indicates the relative geometric complexity of the ROI as compared to the same ROI in the reference plan. Hence GCratio can be used to predict if a defined clinical objective associated with the ROI can be met by the optimizer for a given case. We have evaluated the proposed method on six Head and Neck cases using Pinnacle3 (version 9.10.0) Treatment Planning System (TPS). Out of total of 42 clinical objectives from six cases accounted in the study, 37 were in agreement with the prediction, which implies an agreement of about 88% between predicted and obtained results. The results indicate the feasibility of using the proposed method in head and neck cases for predicting the achievability of clinical objectives.

Sign in / Sign up

Export Citation Format

Share Document