Quantification in-the-wild: data-sets and baselines
Title | Quantification in-the-wild: data-sets and baselines |
Publication Type | Journal Article |
Year of Publication | 2015 |
Authors | Beijbom, O., Hoffman J., Yao E., Darrell T., Rodriguez-Ramirez A., Gonzalez-Rivero M., & Hoegh-Guldberg O. |
Published in | CoRR |
Volume | abs/1510.04811 |
Date Published | 10/2015 |
Abstract | Quantification is the task of estimating the class-distribution of a data-set. While typically considered as a parameter estimation problem with strict assumptions on the data-set shift, we consider quantification in-the-wild, on two large scale data-sets from marine ecology: a survey of Caribbean coral reefs, and a plankton time series from Martha's Vineyard Coastal Observatory. We investigate several quantification methods from the literature and indicate opportunities for future work. In particular, we show that a deep neural network can be fine-tuned on a very limited amount of data (25 - 100 samples) to outperform alternative methods. |
URL | http://www.icsi.berkeley.edu/pubs/vision/quantificationinthewild15.pdf |
ICSI Research Group | Vision |