• Publications
  • Influence
Microsoft COCO: Common Objects in Context
TLDR
We present a new dataset with the goal of advancing the state-of-the-art in object recognition by placing the question of object recognition in the context of the broader question of scene understanding. Expand
  • 11,026
  • 2161
  • PDF
Contour Detection and Hierarchical Image Segmentation
TLDR
We present state-of-the-art algorithms for contour detection and image segmentation. Expand
  • 3,602
  • 631
  • PDF
SVM-KNN: Discriminative Nearest Neighbor Classification for Visual Category Recognition
TLDR
We consider visual category recognition in the framework of measuring similarities, or equivalently perceptual distances, to prototype examples of categories in a homogeneous framework. Expand
  • 1,220
  • 110
  • PDF
FractalNet: Ultra-Deep Neural Networks without Residuals
TLDR
We introduce a design strategy for neural network macro-architecture based on self-similarity. Expand
  • 442
  • 62
  • PDF
Learning Representations for Automatic Colorization
We develop a fully automatic image colorization system. Our approach leverages recent advances in deep networks, exploiting both low-level and semantic representations. As many scene elementsExpand
  • 404
  • 46
  • PDF
From contours to regions: An empirical evaluation
TLDR
We propose a generic grouping algorithm that constructs a hierarchy of regions from the output of any contour detector that produces state-of-the-art image segmentations. Expand
  • 282
  • 42
Using contours to detect and localize junctions in natural images
TLDR
We present a new, high-performance detector for contours in natural images, and use its output to detect and localize image junctions. Expand
  • 459
  • 36
  • PDF
Direct Intrinsics: Learning Albedo-Shading Decomposition by Convolutional Regression
TLDR
We introduce a new approach to intrinsic image decomposition, the task of decomposing a single image into albedo and shading components. Expand
  • 118
  • 26
  • PDF
From contours to regions: An empirical evaluation
TLDR
We propose a generic grouping algorithm that constructs a hierarchy of regions from the output of any contour detector that produces state-of-the-art image segmentations. Expand
  • 222
  • 23
  • PDF
Names and faces in the news
TLDR
We show quite good face clustering is possible for a dataset of inaccurately and ambiguously labelled face images. Expand
  • 262
  • 21