A semi-supervised deep generative model for human body analysis
Deep generative modelling for human body analysis is an emerging problem with many interesting applications. However, the latent space learned by such models is typically not interpretable, resulting in less flexible models. In this work, we adopt a structured semi-supervised approach and present a...
| Main Authors: | , , , , , |
|---|---|
| Format: | Conference item |
| Published: |
Springer, Cham
2019
|
| _version_ | 1826299890096930816 |
|---|---|
| author | De Bem, R Ghosh, A Ajanthan, T Miksik, O Narayanaswamy, S Torr, P |
| author_facet | De Bem, R Ghosh, A Ajanthan, T Miksik, O Narayanaswamy, S Torr, P |
| author_sort | De Bem, R |
| collection | OXFORD |
| description | Deep generative modelling for human body analysis is an emerging problem with many interesting applications. However, the latent space learned by such models is typically not interpretable, resulting in less flexible models. In this work, we adopt a structured semi-supervised approach and present a deep generative model for human body analysis where the body pose and the visual appearance are disentangled in the latent space. Such a disentanglement allows independent manipulation of pose and appearance, and hence enables applications such as pose-transfer without being explicitly trained for such a task. In addition, our setting allows for semi-supervised pose estimation, relaxing the need for labelled data. We demonstrate the capabilities of our generative model on the Human3.6M and on the DeepFashion datasets. |
| first_indexed | 2024-03-07T05:08:49Z |
| format | Conference item |
| id | oxford-uuid:dadc13e6-02e3-4965-b900-4b2b31a332f4 |
| institution | University of Oxford |
| last_indexed | 2024-03-07T05:08:49Z |
| publishDate | 2019 |
| publisher | Springer, Cham |
| record_format | dspace |
| spelling | oxford-uuid:dadc13e6-02e3-4965-b900-4b2b31a332f42022-03-27T09:06:15ZA semi-supervised deep generative model for human body analysisConference itemhttp://purl.org/coar/resource_type/c_5794uuid:dadc13e6-02e3-4965-b900-4b2b31a332f4Symplectic Elements at OxfordSpringer, Cham2019De Bem, RGhosh, AAjanthan, TMiksik, ONarayanaswamy, STorr, PDeep generative modelling for human body analysis is an emerging problem with many interesting applications. However, the latent space learned by such models is typically not interpretable, resulting in less flexible models. In this work, we adopt a structured semi-supervised approach and present a deep generative model for human body analysis where the body pose and the visual appearance are disentangled in the latent space. Such a disentanglement allows independent manipulation of pose and appearance, and hence enables applications such as pose-transfer without being explicitly trained for such a task. In addition, our setting allows for semi-supervised pose estimation, relaxing the need for labelled data. We demonstrate the capabilities of our generative model on the Human3.6M and on the DeepFashion datasets. |
| spellingShingle | De Bem, R Ghosh, A Ajanthan, T Miksik, O Narayanaswamy, S Torr, P A semi-supervised deep generative model for human body analysis |
| title | A semi-supervised deep generative model for human body analysis |
| title_full | A semi-supervised deep generative model for human body analysis |
| title_fullStr | A semi-supervised deep generative model for human body analysis |
| title_full_unstemmed | A semi-supervised deep generative model for human body analysis |
| title_short | A semi-supervised deep generative model for human body analysis |
| title_sort | semi supervised deep generative model for human body analysis |
| work_keys_str_mv | AT debemr asemisuperviseddeepgenerativemodelforhumanbodyanalysis AT ghosha asemisuperviseddeepgenerativemodelforhumanbodyanalysis AT ajanthant asemisuperviseddeepgenerativemodelforhumanbodyanalysis AT miksiko asemisuperviseddeepgenerativemodelforhumanbodyanalysis AT narayanaswamys asemisuperviseddeepgenerativemodelforhumanbodyanalysis AT torrp asemisuperviseddeepgenerativemodelforhumanbodyanalysis AT debemr semisuperviseddeepgenerativemodelforhumanbodyanalysis AT ghosha semisuperviseddeepgenerativemodelforhumanbodyanalysis AT ajanthant semisuperviseddeepgenerativemodelforhumanbodyanalysis AT miksiko semisuperviseddeepgenerativemodelforhumanbodyanalysis AT narayanaswamys semisuperviseddeepgenerativemodelforhumanbodyanalysis AT torrp semisuperviseddeepgenerativemodelforhumanbodyanalysis |