Very Deep Convolutional Networks for Large-Scale Image Recognition
Citations
1,358 citations
Cites background or methods from "Very Deep Convolutional Networks fo..."
...The recent success of deep convolutional neural network (CNN) models [17, 26, 13] has enabled remarkable progress in pixel-wise semantic segmentation tasks due to rich hierarchical features and an end-to-end trainable framework [21, 31, 29, 20, 18, 3]....
[...]
...Significant gains in mean Intersection-overUnion (mIoU) scores on PASCAL VOC2012 dataset [8] were reported when the 16-layer VGG-16 model [26] was replaced by a 101-layer ResNet-101 [13] model [3]; using 152 layer ResNet-152 model yields further improvements [28]....
[...]
...In addition, inspired by the design of the VGG network [26], in that a single 5 × 5 convolutional layer can be decomposed into two adjacent 3×3 convolutional layers to increase the expressiveness of the network while maintaining the receptive field size, we replaced the 7× 7 convolutional layer in the original ResNet-101 network by three 3 × 3 convolutional layers....
[...]
1,342 citations
Cites methods from "Very Deep Convolutional Networks fo..."
...This drastically reduces the memory requirements, enabling four images to fit into the typical GPU memory of 12G. Learning is initialized with the popular VGG-Net [25]....
[...]
...Learning is initialized with the popular VGG-Net [25]....
[...]
1,324 citations
Additional excerpts
...With the advent of deep CNNs, fully connected layers of the network have been commonly employed for image representation [38,43]....
[...]
1,321 citations
1,315 citations
References
49,639 citations
21,729 citations
9,803 citations
9,775 citations
6,061 citations