310 likes | 680 Views
Unbiased Look at Dataset Bias. Antonio Torralba Massachusetts Institute of Technology Alexei A. Efros Carnegie Mellon University CVPR 2011. Outline. 1. Introduction 2. Measuring Dataset Bias 3 . Measuring Dataset’s Value 4 . Discussion. Name That Dataset!.
E N D
Unbiased Look at Dataset Bias Antonio Torralba Massachusetts Institute of Technology Alexei A. Efros Carnegie Mellon University CVPR 2011
Outline • 1. Introduction • 2. Measuring Dataset Bias • 3. Measuring Dataset’s Value • 4. Discussion
Name That Dataset! • Let’s play a game!
UIUC test set is not the same as its training set COILis a lab-baseddataset Caltech101and Caltech256 are predictably confused with each other
Caltech 101 Caltech256 • Pictures of objects belonging to 101 categories. About 40 to 800 images per category • Most categories have about 50images • Collected in September 2003 • The size of each image is roughly 300 x 200 pixels
LabelMe • A project created by the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL) • Adataset of digital images with annotations • The most applicable use of LabelMe is in computer vision research • As of October 31, 2010, LabelMe has 187,240 images, 62,197 annotated images, and 658,992 labeled objects
Bias • Urban scenes Rural landscapes • Professional photographs Amateur snapshots • Entire scenes Single objects
The Rise of the Modern Dataset • COIL-100 dataset (a hundred household objects on a black background) • Coreland 15 Sceneswere Professional collections visual complexity • Caltech-101(101 objects using Google and cleaned by hand) wilderness of the Internet • MSRCand LabelMe(both researcher-collected sets), complex scenes with many objects
The Rise of the Modern Dataset • PASCALVisual Object Classes (VOC) was a reaction against the lax training and testing standards of previous datasets • The batch of very-large-scale, Internet-mined datasets – Tiny Images , ImageNet, and SUN09– can be considered a reaction against the inadequacies of training and testing on datasets that are just too small for the complexity of the real world
Outline • 2. Measuring Dataset Bias -2.1. Cross-dataset generalization -2.2. Negative Set Bias
Negative Set Bias • Evaluate the relative bias in the negative sets of different datasets (e.g. is a “not car” in PASCAL different from “not car” in MSRC?). • For each dataset, we train a classifier on its own set of positive and negative instances. Then, during testing, the positives come from that dataset, but the negativescome from all datasets combined
Outline • 3. Measuring Dataset’s Value
Measuring Dataset’s Value • Given a particular detection task and benchmark, there are two basic ways of improving the performance • The first solution is to improve the features, the object representation and the learning algorithm for the detector • The second solution is to simply enlarge the amount of data available for training
Outline • 4. Discussion
Discussion • Caltech-101is extremely biased with virtually no observed generalization, and should have been retired long ago (as arguedby[14] back in 2006) • MSRChas also fared very poorly. • PASCAL VOC, ImageNet and SUN09, have fared comparatively well