2021 IEEE/CVF International Conference on Computer Vision (ICCV) 2021
DOI: 10.1109/iccv48922.2021.01437
|View full text |Cite
|
Sign up to set email alerts
|

Dressing in Order: Recurrent Person Image Generation for Pose Transfer, Virtual Try-on and Outfit Editing

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1

Citation Types

0
28
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
3
2
2

Relationship

0
7

Authors

Journals

citations
Cited by 44 publications
(29 citation statements)
references
References 30 publications
0
28
0
Order By: Relevance
“…Over the last several years, some works have been dedicated to skeleton-based motion retargeting from 2D inputs. Several works [8,9,19,34] have explored to motion retargeting in image space. Ma et al [19] proposed the pioneer top-down method ๐‘ƒ๐บ 2 , which mainly used conditional GAN [22].…”
Section: Related Workmentioning
confidence: 99%
See 1 more Smart Citation
“…Over the last several years, some works have been dedicated to skeleton-based motion retargeting from 2D inputs. Several works [8,9,19,34] have explored to motion retargeting in image space. Ma et al [19] proposed the pioneer top-down method ๐‘ƒ๐บ 2 , which mainly used conditional GAN [22].…”
Section: Related Workmentioning
confidence: 99%
“…The top-down method usually shows poor results for large differences between source and target poses. To address the above deficiency, based-deformation methods [8,9] are proposed by transferring the feature maps of the source person image into the target pose. To explore the latent feature representation in motion retargeting, Zhao et al [34] extract pose features from an interpolated pose sequence, from source pose to target pose.…”
Section: Related Workmentioning
confidence: 99%
“…ADGAN [30] is incapable of fusing the background and fails to correctly preserve the garment attributes. DiOR [5] is a versatile model that can handle The first two columns represent the inputs, while the others are garment transfer results from our method and the other three baselines (LWG [27], ADGAN [30] and DiOR [5]). Our wFlow contains richer foreground and background texture details and more successfully transfer the loose garments.…”
Section: Comparison With Baselinesmentioning
confidence: 99%
“…Figure 4. Qualitative comparisons on Dance50k (1-3rd rows) and DeepFashion Datset (4-5th rows).The first two columns represent the inputs, while the others are garment transfer results from our method and the other three baselines (LWG[27], ADGAN[30] and DiOR[5]). Our wFlow contains richer foreground and background texture details and more successfully transfer the loose garments.…”
mentioning
confidence: 99%
See 1 more Smart Citation