Výsledky vyhledávání - "Kim, Taeksoo"

Report

GALA: Generating Animatable Layered Assets from a Single Scan

Autor: Kim, Taeksoo, Kim, Byungjun, Saito, Shunsuke, Joo, Hanbyul

We present GALA, a framework that takes as input a single-layer clothed 3D human mesh and decomposes it into complete multi-layered 3D assets. The outputs can then be combined with other assets to create novel clothed human avatars with any pose. Exi

Externí odkaz: http://arxiv.org/abs/2401.12979

Zobrazit plný text záznamu

Report

NCHO: Unsupervised Learning for Neural 3D Composition of Humans and Objects

Autor: Kim, Taeksoo, Saito, Shunsuke, Joo, Hanbyul

Deep generative models have been recently extended to synthesizing 3D digital humans. However, previous approaches treat clothed humans as a single chunk of geometry without considering the compositionality of clothing and accessories. As a result, i

Externí odkaz: http://arxiv.org/abs/2305.14345

Zobrazit plný text záznamu

Report

Semi-supervised Feature-Level Attribute Manipulation for Fashion Image Retrieval

Autor: Shin, Minchul, Park, Sanghyuk, Kim, Taeksoo

With a growing demand for the search by image, many works have studied the task of fashion instance-level image retrieval (FIR). Furthermore, the recent works introduce a concept of fashion attribute manipulation (FAM) which manipulates a specific at

Externí odkaz: http://arxiv.org/abs/1907.05007

Zobrazit plný text záznamu

Report

Unsupervised Visual Attribute Transfer with Reconfigurable Generative Adversarial Networks

Autor: Kim, Taeksoo, Kim, Byoungjip, Cha, Moonsu, Kim, Jiwon

Learning to transfer visual attributes requires supervision dataset. Corresponding images with varying attribute values with the same identity are required for learning the transfer function. This largely limits their applications, because capturing

Externí odkaz: http://arxiv.org/abs/1707.09798

Zobrazit plný text záznamu

Report

Learning to Discover Cross-Domain Relations with Generative Adversarial Networks

Autor: Kim, Taeksoo, Cha, Moonsu, Kim, Hyunsoo, Lee, Jung Kwon, Kim, Jiwon

While humans easily recognize relations between data from different domains without any supervision, learning to automatically discover them is in general very challenging and needs many ground-truth pairs that illustrate the relations. To avoid cost

Externí odkaz: http://arxiv.org/abs/1703.05192

Zobrazit plný text záznamu