DMCA
Fast, Accurate Detection of 100,000 Object Classes on a Single Machine
Cached
Download Links
| Citations: | 33 - 1 self |
BibTeX
@MISC{Dean_fast,accurate,
author = {Thomas Dean and Mark A. Ruzon and Mark Segal and Jonathon Shlens and Sudheendra Vijayanarasimhan and Jay Yagnik},
title = {Fast, Accurate Detection of 100,000 Object Classes on a Single Machine},
year = {}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Many object detection systems are constrained by the time required to convolve a target image with a bank of filters that code for different aspects of an object’s appearance, such as the presence of component parts. We exploit locality-sensitive hashing to replace the dot-product kernel operator in the convolution with a fixed number of hash-table probes that effectively sample all of the filter responses in time independent of the size of the filter bank. To show the effectiveness of the technique, we apply it to evaluate 100,000 deformable-part models requiring over a million (part) filters on multiple scales of a target image in less than 20 seconds using a single multi-core processor with 20GB of RAM. This represents a speed-up of approximately 20,000 times — four orders of magnitude — when compared with performing the convolutions explicitly on the same hardware. While mean average precision over the full set of 100,000 object classes is around 0.16 due in large part to the challenges in gathering training data and collecting ground truth for so many classes, we achieve a mAP of at least 0.20 on a third of the classes and 0.30 or better on about 20 % of the classes. 1.
Keyphrases
object class single machine accurate detection target image mean average precision many object detection system dot-product kernel operator deformable-part model object appearance full set fixed number filter bank different aspect ground truth training data large part hash-table probe component part filter response many class single multi-core processor multiple scale locality-sensitive hashing
