Abstract

Recognition algorithms based on convolutional networks (CNNs) typically use the output of the last layer as a feature representation. However, the information in this layer may be too coarse spatially to allow precise localization. On the contrary, earlier layers may be precise in localization but will not capture semantics. To get the best of both worlds, we define the hypercolumn at a pixel as the vector of activations of all CNN units above that pixel. Using hypercolumns as pixel descriptors, we show results on three fine-grained localization tasks: simultaneous detection and segmentation [22], where we improve state-of-the-art from 49.7 mean AP <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">r</sup> [22] to 60.0, keypoint localization, where we get a 3.3 point boost over [20], and part labeling, where we show a 6.6 point gain over a strong baseline.

Keywords

PixelArtificial intelligenceSegmentationComputer scienceRepresentation (politics)Pattern recognition (psychology)Convolutional neural networkSemantics (computer science)Feature (linguistics)Object (grammar)Point (geometry)Layer (electronics)Computer visionMathematics

Affiliated Institutions

Related Publications

Compressed sensing

Suppose x is an unknown vector in Ropf <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">m</sup> (a digital image or signal); we pla...

2006 IEEE Transactions on Information Theory 22524 citations

Publication Info

Year
2015
Type
article
Pages
447-456
Citations
1577
Access
Closed

External Links

Social Impact

Social media, news, blog, policy document mentions

Citation Metrics

1577
OpenAlex

Cite This

Bharath Hariharan, Pablo Arbeláez, Ross Girshick et al. (2015). Hypercolumns for object segmentation and fine-grained localization. , 447-456. https://doi.org/10.1109/cvpr.2015.7298642

Identifiers

DOI
10.1109/cvpr.2015.7298642