AI Paper Atlas

Research map / 3D vision / 3D generation

WordRobe: Text-Guided Generation of Textured 3D Garments

Astitva Srivastava, Pranav Manu, Amit Raj, Varun Jampani, Avinash Sharma

ECCV 2024 · 3D vision › 3D generation · approach: Text-to-3D scene generation · 16 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use two-stage coarse-to-fine decoding instead of single decoder (ablation)WordRobe garment encoder-decoder · 3D garment dataset (unseen categories)Chamfer Distance: 1.631 -> 1.078Improved
add gradient loss L_grad (ablation)WordRobe garment encoder-decoder · 3D garment dataset (unseen categories)Chamfer Distance: 1.886 -> 1.078Improved
add latent disentanglement loss L_latent (ablation)WordRobe garment encoder-decoder · 3D garment dataset (unseen categories)Chamfer Distance: 1.094 -> 1.078No clear effect
add latent disentanglement loss L_latent (ablation)WordRobe garment encoder-decoder · Latent interpolationdelta_vol: 1.275 -> 1.206Improved
train mapping network with L1 loss instead of L2 or L1+cosine (ablation)MLP_map · Test garmentsMSE: 0.7433 (L2) -> 0.3481Improved
composite front/back depth maps into one ControlNet pass instead of multiview optimization (Text2Tex)WordRobe texture synthesis · Text-guided texture synthesisCLIP Score (ViT-H/14): 0.263 -> 0.304Improved
use view-composited ControlNet texture synthesis instead of Text2TexWordRobe texture synthesis · Texture synthesis runtimetime per prompt: 5 min -> 22 secondsImproved
train WordRobe on its own garment dataset and test on CLOTH3D versus DrapeNet trained on CLOTH3DWordRobe garment encoder-decoder · CLOTH3D topwearP2S: 0.631 vs 0.635No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy