Towards Garment Sewing Pattern Reconstruction from a Single Image

Published in

ACM Transactions on Graphics(2023)

External Links:

Generate Graph DownloadPDF

TL;DR

This work synthesizes a versatile dataset, named SewFactory, which consists of around 1M images and ground-truth sewing patterns for model training and quantitative evaluation, and proposes a two-level Transformer network called Sewformer, which significantly improves the sewing pattern prediction performance.

Abstract

Garment sewing pattern represents the intrinsic rest shape of a garment, and is the core for many applications like fashion design, virtual try-on, and digital avatars. In this work, we explore the challenging problem of recovering garment sewing patterns from daily photos for augmenting these applications. To solve the problem, we first synthesize a versatile dataset, named SewFactory, which consists of around 1M images and ground-truth sewing patterns for model training and quantitative evaluation. SewFactory covers a wide range of human poses, body shapes, and sewing patterns, and possesses realistic appearances thanks to the proposed human texture synthesis network. Then, we propose a two-level Transformer network called Sewformer, which significantly improves the sewing pattern prediction performance. Extensive experiments demonstrate that the proposed framework is effective in recovering sewing patterns and well generalizes to casually-taken human photos. Code, dataset, and pre-trained models will be released.

Authors

Shuicheng Yan

2 papers

Lijuan Liu

1 papers

Xiangyu Xu

1 papers

References72 items

SMPL: A Skinned Multi-Person Linear Model

ECON: Explicit Clothed humans Optimized via Normal integration

Structure-Preserving 3D Garment Modeling with Neural Sewing Machines

3D Clothed Human Reconstruction in the Wild

Dressing Avatars

Towards Garment Sewing Pattern Reconstruction from a Single Image

Published in

ACM Transactions on Graphics(2023)

External Links:

Generate Graph DownloadPDF

TL;DR

Abstract

Authors

Shuicheng Yan

2 papers

Lijuan Liu

1 papers

Xiangyu Xu

1 papers

References72 items

SMPL: A Skinned Multi-Person Linear Model

ECON: Explicit Clothed humans Optimized via Normal integration

Structure-Preserving 3D Garment Modeling with Neural Sewing Machines

3D Clothed Human Reconstruction in the Wild

Dressing Avatars

Jiabin Liang

1 papers

StyleGAN-Human: A Data-Centric Odyssey of Human Generation

AUV-Net: Learning Aligned UV Maps for Texture Transfer and Synthesis

Generative Adversarial Networks

DeepDraper: Fast and Accurate 3D Garment Draping over a 3D Human Body

Generating Datasets of 3D Garments with Sewing Patterns

3D Human Texture Estimation from a Single Image with Transformers

Learning Anchored Unsigned Distance Functions with Gradient Direction Alignment for Single-view Garment Reconstruction

SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations

Estimating Garment Patterns from Static Scan Data

TL;DR

Abstract

Authors

References72 items

SMPL: A Skinned Multi-Person Linear Model

ECON: Explicit Clothed humans Optimized via Normal integration

Structure-Preserving 3D Garment Modeling with Neural Sewing Machines

3D Clothed Human Reconstruction in the Wild

Dressing Avatars

TL;DR

Abstract

Authors

References72 items

SMPL: A Skinned Multi-Person Linear Model

ECON: Explicit Clothed humans Optimized via Normal integration

Structure-Preserving 3D Garment Modeling with Neural Sewing Machines

3D Clothed Human Reconstruction in the Wild

Dressing Avatars

StyleGAN-Human: A Data-Centric Odyssey of Human Generation

AUV-Net: Learning Aligned UV Maps for Texture Transfer and Synthesis

Generative Adversarial Networks

DeepDraper: Fast and Accurate 3D Garment Draping over a 3D Human Body

Pose with style

Generating Datasets of 3D Garments with Sewing Patterns

3D Human Texture Estimation from a Single Image with Transformers

Learning Anchored Unsigned Distance Functions with Gradient Direction Alignment for Single-view Garment Reconstruction

SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations

Estimating Garment Patterns from Static Scan Data

StylePeople: A Generative Model of Fullbody Human Avatars

Semi-supervised Synthesis of High-Resolution Editable Textures for 3D Humans

SMPLicit: Topology-aware Generative Model for Clothed People

HumanGAN: A Generative Model of Human Images

Style and Pose Control for Image Synthesis of Humans from a Single Monocular View

Neural Re-rendering of Humans from a Single Image

3D Human Shape and Pose from a Single Low-Resolution Image with Self-Supervised Learning

SIZER: A Dataset and Model for Parsing 3D Clothing and Learning Size Sensitive 3D Clothing

MISC: Multi-Condition Injection and Spatially-Adaptive Compositing for Conditional Person Image Synthesis

Cross-Domain Correspondence Learning for Exemplar-Based Image Translation

BCNet: Learning Body and Cloth Shape from A Single Image

PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization

Deep Fashion3D: A Dataset and Benchmark for 3D Garment Reconstruction from Single Images

CLOTH3D: Clothed 3D Humans

ClothFlow: A Flow-Based Model for Clothed Person Generation

360-Degree Textures of People in Clothing from a Single Image

Multi-Garment Net: Learning to Dress 3D People From Images

Learning to Dress 3D People in Generative Clothing

3DPeople: Modeling the Geometry of Dressed Humans

AMASS: Archive of Motion Capture As Surface Shapes

A Style-Based Generator Architecture for Generative Adversarial Networks

Physics-Inspired Garment Recovery from a Single-View Image

3D interactive garment pattern-making technology

Arnold

Variational surface cutting

Learning a shared shape space for multimodal garment design

A Generative Model of People in Clothing

Detailed, Accurate, Human Shape Estimation from Clothed 3D Scan Sequences

Learning from Synthetic Humans

Image-to-Image Translation with Conditional Adversarial Networks

DeepFashion: Powering Robust Clothes Recognition and Retrieval with Rich Annotations

Perceptual Losses for Real-Time Style Transfer and Super-Resolution

Deep Residual Learning for Image Recognition

Garment modeling with a depth camera

U-Net: Convolutional Networks for Biomedical Image Segmentation

Garment capture from a photograph

Adaptive anisotropic remeshing for cloth simulation

Reverse Engineering Garments

Interactive decal compositing with discrete exponential maps

Adaptive filtering via particle swarm optimization

Stable but responsive cloth

Data-driven Garment Pattern Estimation from 3D Geometries

Swin Transformer: Hierarchical Vision Transformer using Shifted Windows

Human Parsing Based Texture Transfer from Single Image to 3D Human via Cross-View Consistency

Tailornet:Predictingclothingin3dasafunctionofhumanpose,shapeandgarmentstyle

Arch: Ani-matablereconstructionofclothedhumans

LearningtoreconstructpeopleinclothingfromasingleRGBcamera

Example-guidedstyle-consistentimagesynthesisfromsemantic labeling

Pixel2mesh:Generating3dmeshmodelsfromsinglergbimages

Texturegan: Controlling deep image synthesis

Attentionisallyouneed

Learningtosuper-resolveblurryfaceandtextimages