Utilizing object-object and object-scene context when planning to find things
In this paper, our goal is to search for a novel object, where we have a prior map of the environment and knowledge of some of the objects in it, but no information about the location of the specific novel object. We develop a probabilistic model over possible object locations that utilizes object-o...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | en_US |
| Published: |
Institute of Electrical and Electronics Engineers
2011
|
| Online Access: | http://hdl.handle.net/1721.1/66603 https://orcid.org/0000-0002-8293-0492 |
| _version_ | 1826208443584741376 |
|---|---|
| author | Kollar, Thomas Fleming Roy, Nicholas |
| author2 | Massachusetts Institute of Technology. Department of Aeronautics and Astronautics |
| author_facet | Massachusetts Institute of Technology. Department of Aeronautics and Astronautics Kollar, Thomas Fleming Roy, Nicholas |
| author_sort | Kollar, Thomas Fleming |
| collection | MIT |
| description | In this paper, our goal is to search for a novel object, where we have a prior map of the environment and knowledge of some of the objects in it, but no information about the location of the specific novel object. We develop a probabilistic model over possible object locations that utilizes object-object and object-scene context. This model can be queried for any of over 25,000 naturally occurring objects in the world and is trained from labeled data acquired from the captions of photos on the Flickr Website. We show that these simple models based on object co-occurrences perform surprisingly well at localizing arbitrary objects in an office setting. In addition, we show how to compute paths that minimize the expected distance to the query object and show that this approach performs better than a greedy approach. Finally, we give preliminary results for grounding our approach in object classifiers. |
| first_indexed | 2024-09-23T14:05:50Z |
| format | Article |
| id | mit-1721.1/66603 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T14:05:50Z |
| publishDate | 2011 |
| publisher | Institute of Electrical and Electronics Engineers |
| record_format | dspace |
| spelling | mit-1721.1/666032022-10-01T19:10:19Z Utilizing object-object and object-scene context when planning to find things Kollar, Thomas Fleming Roy, Nicholas Massachusetts Institute of Technology. Department of Aeronautics and Astronautics Roy, Nicholas Kollar, Thomas Fleming Roy, Nicholas In this paper, our goal is to search for a novel object, where we have a prior map of the environment and knowledge of some of the objects in it, but no information about the location of the specific novel object. We develop a probabilistic model over possible object locations that utilizes object-object and object-scene context. This model can be queried for any of over 25,000 naturally occurring objects in the world and is trained from labeled data acquired from the captions of photos on the Flickr Website. We show that these simple models based on object co-occurrences perform surprisingly well at localizing arbitrary objects in an office setting. In addition, we show how to compute paths that minimize the expected distance to the query object and show that this approach performs better than a greedy approach. Finally, we give preliminary results for grounding our approach in object classifiers. United States. Office of Naval Research. Multidisciplinary University Research Initiative (N00014-07-1-0749) 2011-10-27T18:24:20Z 2011-10-27T18:24:20Z 2009-07 Article http://purl.org/eprint/type/ConferencePaper 978-1-4244-2788-8 http://hdl.handle.net/1721.1/66603 Kollar, T., and N. Roy. “Utilizing object-object and object-scene context when planning to find things.” IEEE, 2009. 2168-2173. Web. 27 Oct. 2011. © 2009 Institute of Electrical and Electronics Engineers https://orcid.org/0000-0002-8293-0492 en_US http://dx.doi.org/10.1109/ROBOT.2009.5152831 IEEE International Conference on Robotics and Automation, 2009 Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use. application/pdf Institute of Electrical and Electronics Engineers IEEE |
| spellingShingle | Kollar, Thomas Fleming Roy, Nicholas Utilizing object-object and object-scene context when planning to find things |
| title | Utilizing object-object and object-scene context when planning to find things |
| title_full | Utilizing object-object and object-scene context when planning to find things |
| title_fullStr | Utilizing object-object and object-scene context when planning to find things |
| title_full_unstemmed | Utilizing object-object and object-scene context when planning to find things |
| title_short | Utilizing object-object and object-scene context when planning to find things |
| title_sort | utilizing object object and object scene context when planning to find things |
| url | http://hdl.handle.net/1721.1/66603 https://orcid.org/0000-0002-8293-0492 |
| work_keys_str_mv | AT kollarthomasfleming utilizingobjectobjectandobjectscenecontextwhenplanningtofindthings AT roynicholas utilizingobjectobjectandobjectscenecontextwhenplanningtofindthings |