Fully convolutional networks for semantic segmentation
Citations
123,388 citations
49,590 citations
44,703 citations
27,821 citations
26,458 citations
References
3,945 citations
"Fully convolutional networks for se..." refers background in this paper
...[19] discard the nonconvolutional portion of classification nets to make a feature extractor....
[...]
...These include advances in bounding box object detection [32, 12, 19], part and keypoint prediction [42, 26], and local correspondence [26, 10]....
[...]
3,760 citations
3,043 citations
2,994 citations
2,791 citations
"Fully convolutional networks for se..." refers background or methods or result in this paper
...In contrast, previous works have applied small convnets without supervised pre-training [9, 31, 30]....
[...]
...Our approach does not make use of pre- and post-processing complications, including superpixels [9, 17], proposals [17, 15], or post-hoc refinement by random fields or local classifiers [9, 17]....
[...]
...2 Sampling in patchwise training can correct class imbalance [30, 9, 3] and mitigate the spatial correlation of dense patches [31, 17]....
[...]
...Common elements of these approaches include • small models restricting capacity and receptive fields; • patchwise training [30, 3, 9, 31, 11]; • post-processing by superpixel projection, random field regularization, filtering, or local classification [9, 3, 11]; • input shifting and output interlacing for dense output [32, 31, 11]; • multi-scale pyramid processing [9, 31, 11]; • saturating tanh nonlinearities [9, 6, 31]; and • ensembles [3, 11], whereas our method does without this machinery....
[...]
...Prior approaches have used convnets for semantic segmentation [30, 3, 9, 31, 17, 15, 11], in which each pixel is labeled with the class of its enclosing object or region, but with shortcomings that this work addresses....
[...]