Video retrieval by mimicking poses
We describe a method for real time video retrieval where the task is to match the 2D human pose of a query. A user can form a query by (i) interactively controlling a stickman on a web based GUI, (ii) uploading an image of the desired pose, or (iii) using the Kinect and acting out the query himself....
| Main Authors: | , , , , |
|---|---|
| Format: | Conference item |
| Language: | English |
| Published: |
Association for Computing Machinery
2012
|
| _version_ | 1817932410686275584 |
|---|---|
| author | Jammalamadaka, N Zisserman, A Eichner, M Ferrari, V Jawahar, CV |
| author_facet | Jammalamadaka, N Zisserman, A Eichner, M Ferrari, V Jawahar, CV |
| author_sort | Jammalamadaka, N |
| collection | OXFORD |
| description | We describe a method for real time video retrieval where the task is to match the 2D human pose of a query. A user can form a query by (i) interactively controlling a stickman on a web based GUI, (ii) uploading an image of the desired pose, or (iii) using the Kinect and acting out the query himself. The method is scalable and is applied to a dataset of 18 films totaling more than three million frames. The real time performance is achieved by searching for approximate nearest neighbors to the query using a random forest of K-D trees. Apart from the query modalities, we introduce two other areas of novelty. First, we show that pose retrieval can proceed using a low dimensional representation. Second, we show that the precision of the results can be improved substantially by combining the outputs of independent human pose estimation algorithms. The performance of the system is assessed quantitatively over a range of pose queries. |
| first_indexed | 2024-12-09T03:37:29Z |
| format | Conference item |
| id | oxford-uuid:8373677c-60cf-4bae-aa07-c0fb3a258398 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-12-09T03:37:29Z |
| publishDate | 2012 |
| publisher | Association for Computing Machinery |
| record_format | dspace |
| spelling | oxford-uuid:8373677c-60cf-4bae-aa07-c0fb3a2583982024-12-03T14:36:02ZVideo retrieval by mimicking posesConference itemhttp://purl.org/coar/resource_type/c_5794uuid:8373677c-60cf-4bae-aa07-c0fb3a258398EnglishSymplectic ElementsAssociation for Computing Machinery2012Jammalamadaka, NZisserman, AEichner, MFerrari, VJawahar, CVWe describe a method for real time video retrieval where the task is to match the 2D human pose of a query. A user can form a query by (i) interactively controlling a stickman on a web based GUI, (ii) uploading an image of the desired pose, or (iii) using the Kinect and acting out the query himself. The method is scalable and is applied to a dataset of 18 films totaling more than three million frames. The real time performance is achieved by searching for approximate nearest neighbors to the query using a random forest of K-D trees. Apart from the query modalities, we introduce two other areas of novelty. First, we show that pose retrieval can proceed using a low dimensional representation. Second, we show that the precision of the results can be improved substantially by combining the outputs of independent human pose estimation algorithms. The performance of the system is assessed quantitatively over a range of pose queries. |
| spellingShingle | Jammalamadaka, N Zisserman, A Eichner, M Ferrari, V Jawahar, CV Video retrieval by mimicking poses |
| title | Video retrieval by mimicking poses |
| title_full | Video retrieval by mimicking poses |
| title_fullStr | Video retrieval by mimicking poses |
| title_full_unstemmed | Video retrieval by mimicking poses |
| title_short | Video retrieval by mimicking poses |
| title_sort | video retrieval by mimicking poses |
| work_keys_str_mv | AT jammalamadakan videoretrievalbymimickingposes AT zissermana videoretrievalbymimickingposes AT eichnerm videoretrievalbymimickingposes AT ferrariv videoretrievalbymimickingposes AT jawaharcv videoretrievalbymimickingposes |