Batch-Mode Active Learning for Traffic Sign Recognition. Nienhüser, D. & Zöllner, J. M. In IEEE Intelligent Vehicles Symposium, 2013.
Batch-Mode Active Learning for Traffic Sign Recognition [link]Mendeley  abstract   bibtex   
Cognitive vehicles face a huge variety of different objects when perceiving their environment using video cameras. Their capability of recognizing certain types of objects in a robust way is tightly coupled to successfully discriminating positive samples from negative ones (background objects). Machine learning methods trained on huge training sets can accomplish this, but to reduce labeling costs and classification runtime it is desirable to minimize the number of training samples needed. We examine the feasibility of active learning paradigms to achieve this goal and propose a batch-mode multi-class active learning query strategy for SVM: Angular diversity ranking is a weighted combination of relevance of unlabeled samples with a diversity measure. Experiments using our query strategy for traffic sign recognition, track-pruned class-biased angular diversity ranking, show a significant reduction of the number of false positives and labeling effort on a large data set.
@inproceedings{ mendeley_5149961331,
  isauthor = {1},
  abstract = {Cognitive vehicles face a huge variety of different objects when perceiving their environment using video cameras. Their capability of recognizing certain types of objects in a robust way is tightly coupled to successfully discriminating positive samples from negative ones (background objects). Machine learning methods trained on huge training sets can accomplish this, but to reduce labeling costs and classification runtime it is desirable to minimize the number of training samples needed. We examine the feasibility of active learning paradigms to achieve this goal and propose a batch-mode multi-class active learning query strategy for SVM: Angular diversity ranking is a weighted combination of relevance of unlabeled samples with a diversity measure. Experiments using our query strategy for traffic sign recognition, track-pruned class-biased angular diversity ranking, show a significant reduction of the number of false positives and labeling effort on a large data set.},
  canonical_id = {13a574a3-5626-3696-b2cf-394149eae0a7},
  added = {1357810856},
  year = {2013},
  keywords = {self},
  isstarred = {0},
  id = {5149961331},
  discipline = {Computer and Information Science},
  booktitle = {IEEE Intelligent Vehicles Symposium},
  title = {Batch-Mode Active Learning for Traffic Sign Recognition},
  deletionpending = {0},
  version = {1380308292},
  folders_ids = {17744653, 51364911},
  type = {Conference Proceedings},
  url_mendeley = {http://www.mendeley.com/research/batchmode-active-learning-traffic-sign-recognition/},
  tags = {active-learning},
  isread = {0},
  author = {Dennis {Nienhüser} and J. Marius {Zöllner}},
  series = {IEEE Intelligent Vehicles Symposium},
  modified = {1380308292},
  citation_key = {Nienhuser2013},
  subdiscipline = {Artificial Intelligence}
}

Downloads: 0