Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning

arxiv.orgsite

For improving GoogLeNet with residual connections.

similar

Inceptionism: Going Deeper into Neural Networks@Google Research googleresearch.blogspot.kr
site category #deep-vision link ↗
Going Deeper with Convolutions cv-foundation.org
site category #deep-learning-resources link ↗

GoogLeNet: Appearance of "Inception" layers/modules, the idea is of parallelizing conv layers into many mini-conv of different size with "same" padding, concatenated on depth.
Visualizing and Understanding Convolutional Networks arxiv.org
site category #deep-learning-resources link ↗

For the "deconvnet layer".
Improving Inception and Image Classification in TensorFlow research.googleblog.com
site category #deep-learning-resources link ↗

Very interesting CNN architecture (e.g.: the inception-style convolutional layers is promising and efficient in terms of reducing the number of parameters).
Very Deep Convolutional Networks for Large-Scale Image Recognition arxiv.org
site category #deep-learning-resources link ↗

Interesting idea of stacking multiple 3x3 conv+ReLU before pooling for a bigger filter size with just a few parameters. There is also a nice table for "ConvNet Configuration".
Individual differences among deep neural network models kriegeskortelab.zuckermaninstitute.columbia.edu
site category #agi-cocosci link ↗

Nature Communications, 2020. [All Versions].

Education › Papers > Convolutional Neural Networks: “For improving GoogLeNet with residual connections.”