Evaluation of image features using a photorealistic virtual world
Image features are widely used in computer vision applications. They need to be robust to scene changes and image transformations. Designing and comparing feature descriptors requires the ability to evaluate their performance with respect to those transformations. We want to know how robust the desc...
| Main Authors: | , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
Institute of Electrical and Electronics Engineers (IEEE)
2012
|
| Online Access: | http://hdl.handle.net/1721.1/72588 https://orcid.org/0000-0002-2231-7995 https://orcid.org/0000-0003-4915-0256 |
| _version_ | 1826197525028143104 |
|---|---|
| author | Kaneva, Biliana K. Torralba, Antonio Freeman, William T. |
| author2 | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory |
| author_facet | Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Kaneva, Biliana K. Torralba, Antonio Freeman, William T. |
| author_sort | Kaneva, Biliana K. |
| collection | MIT |
| description | Image features are widely used in computer vision applications. They need to be robust to scene changes and image transformations. Designing and comparing feature descriptors requires the ability to evaluate their performance with respect to those transformations. We want to know how robust the descriptors are to changes in the lighting, scene, or viewing conditions. For this, we need ground truth data of different scenes viewed under different camera or lighting conditions in a controlled way. Such data is very difficult to gather in a real-world setting. We propose using a photorealistic virtual world to gain complete and repeatable control of the environment in order to evaluate image features. We calibrate our virtual world evaluations by comparing against feature rankings made from photographic data of the same subject matter (the Statue of Liberty). We find very similar feature rankings between the two datasets. We then use our virtual world to study the effects on descriptor performance of controlled changes in viewpoint and illumination. We also study the effect of augmenting the descriptors with depth information to improve performance. |
| first_indexed | 2024-09-23T10:48:59Z |
| format | Article |
| id | mit-1721.1/72588 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T10:48:59Z |
| publishDate | 2012 |
| publisher | Institute of Electrical and Electronics Engineers (IEEE) |
| record_format | dspace |
| spelling | mit-1721.1/725882022-09-27T15:10:57Z Evaluation of image features using a photorealistic virtual world Kaneva, Biliana K. Torralba, Antonio Freeman, William T. Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science Freeman, William T. Kaneva, Biliana K. Torralba, Antonio Freeman, William T. Image features are widely used in computer vision applications. They need to be robust to scene changes and image transformations. Designing and comparing feature descriptors requires the ability to evaluate their performance with respect to those transformations. We want to know how robust the descriptors are to changes in the lighting, scene, or viewing conditions. For this, we need ground truth data of different scenes viewed under different camera or lighting conditions in a controlled way. Such data is very difficult to gather in a real-world setting. We propose using a photorealistic virtual world to gain complete and repeatable control of the environment in order to evaluate image features. We calibrate our virtual world evaluations by comparing against feature rankings made from photographic data of the same subject matter (the Statue of Liberty). We find very similar feature rankings between the two datasets. We then use our virtual world to study the effects on descriptor performance of controlled changes in viewpoint and illumination. We also study the effect of augmenting the descriptors with depth information to improve performance. Quanta Computer (Firm) Shell Research United States. Office of Naval Research. Multidisciplinary University Research Initiative (Grant N00014-06-1-0734) United States. Office of Naval Research. Multidisciplinary University Research Initiative. CAREER (Award Number 0747120) United States. Office of Naval Research. Multidisciplinary University Research Initiative (Grant N000141010933) Microsoft Corporation Adobe Systems Google (Firm) 2012-09-10T14:31:46Z 2012-09-10T14:31:46Z 2011-11 2011-11 Article http://purl.org/eprint/type/ConferencePaper 978-1-4577-1101-5 1550-5499 http://hdl.handle.net/1721.1/72588 Kaneva, Biliana, Antonio Torralba, and William T. Freeman. “Evaluation of Image Features Using a Photorealistic Virtual World.” IEEE International Conference on Computer Vision 2011 (ICCV). 2282–2289. https://orcid.org/0000-0002-2231-7995 https://orcid.org/0000-0003-4915-0256 en_US http://dx.doi.org/10.1109/ICCV.2011.6126508 IEEE International Conference on Computer Vision 2011 (ICCV) Creative Commons Attribution-Noncommercial-Share Alike 3.0 http://creativecommons.org/licenses/by-nc-sa/3.0/ application/pdf Institute of Electrical and Electronics Engineers (IEEE) MIT web domain |
| spellingShingle | Kaneva, Biliana K. Torralba, Antonio Freeman, William T. Evaluation of image features using a photorealistic virtual world |
| title | Evaluation of image features using a photorealistic virtual world |
| title_full | Evaluation of image features using a photorealistic virtual world |
| title_fullStr | Evaluation of image features using a photorealistic virtual world |
| title_full_unstemmed | Evaluation of image features using a photorealistic virtual world |
| title_short | Evaluation of image features using a photorealistic virtual world |
| title_sort | evaluation of image features using a photorealistic virtual world |
| url | http://hdl.handle.net/1721.1/72588 https://orcid.org/0000-0002-2231-7995 https://orcid.org/0000-0003-4915-0256 |
| work_keys_str_mv | AT kanevabilianak evaluationofimagefeaturesusingaphotorealisticvirtualworld AT torralbaantonio evaluationofimagefeaturesusingaphotorealisticvirtualworld AT freemanwilliamt evaluationofimagefeaturesusingaphotorealisticvirtualworld |