An Analysis of the Effect of Gaussian Error in Object Recognition
Object recognition is complicated by clutter, occlusion, and sensor error. Since pose hypotheses are based on image feature locations, these effects can lead to false negatives and positives. In a typical recognition algorithm, pose hypotheses are tested against the image, and a score is assi...
| Main Author: | |
|---|---|
| Language: | en_US |
| Published: |
2004
|
| Online Access: | http://hdl.handle.net/1721.1/7057 |
| _version_ | 1826205791946801152 |
|---|---|
| author | Sarachik, Karen Beth |
| author_facet | Sarachik, Karen Beth |
| author_sort | Sarachik, Karen Beth |
| collection | MIT |
| description | Object recognition is complicated by clutter, occlusion, and sensor error. Since pose hypotheses are based on image feature locations, these effects can lead to false negatives and positives. In a typical recognition algorithm, pose hypotheses are tested against the image, and a score is assigned to each hypothesis. We use a statistical model to determine the score distribution associated with correct and incorrect pose hypotheses, and use binary hypothesis testing techniques to distinguish between them. Using this approach we can compare algorithms and noise models, and automatically choose values for internal system thresholds to minimize the probability of making a mistake. |
| first_indexed | 2024-09-23T13:19:11Z |
| id | mit-1721.1/7057 |
| institution | Massachusetts Institute of Technology |
| language | en_US |
| last_indexed | 2024-09-23T13:19:11Z |
| publishDate | 2004 |
| record_format | dspace |
| spelling | mit-1721.1/70572019-04-10T11:52:29Z An Analysis of the Effect of Gaussian Error in Object Recognition Sarachik, Karen Beth Object recognition is complicated by clutter, occlusion, and sensor error. Since pose hypotheses are based on image feature locations, these effects can lead to false negatives and positives. In a typical recognition algorithm, pose hypotheses are tested against the image, and a score is assigned to each hypothesis. We use a statistical model to determine the score distribution associated with correct and incorrect pose hypotheses, and use binary hypothesis testing techniques to distinguish between them. Using this approach we can compare algorithms and noise models, and automatically choose values for internal system thresholds to minimize the probability of making a mistake. 2004-10-20T20:24:12Z 2004-10-20T20:24:12Z 1994-02-01 AITR-1469 http://hdl.handle.net/1721.1/7057 en_US AITR-1469 7376380 bytes 3521585 bytes application/postscript application/pdf application/postscript application/pdf |
| spellingShingle | Sarachik, Karen Beth An Analysis of the Effect of Gaussian Error in Object Recognition |
| title | An Analysis of the Effect of Gaussian Error in Object Recognition |
| title_full | An Analysis of the Effect of Gaussian Error in Object Recognition |
| title_fullStr | An Analysis of the Effect of Gaussian Error in Object Recognition |
| title_full_unstemmed | An Analysis of the Effect of Gaussian Error in Object Recognition |
| title_short | An Analysis of the Effect of Gaussian Error in Object Recognition |
| title_sort | analysis of the effect of gaussian error in object recognition |
| url | http://hdl.handle.net/1721.1/7057 |
| work_keys_str_mv | AT sarachikkarenbeth ananalysisoftheeffectofgaussianerrorinobjectrecognition AT sarachikkarenbeth analysisoftheeffectofgaussianerrorinobjectrecognition |