- CAM: Learning Deep Features for Discriminative Localization
- Grad-CAM: Grad-CAM: Why did you say that? Visual Explanations from Deep Networks via Gradient-based Localization
- All Convolutional Net (introduced Guided Backpropagation): Striving for Simplicity: The All Convolutional Net
- Deconvnet / Occlusion sensitivity analysis: Visualizing and Understanding Convolutional Networks
- Saliency Maps: Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps (2014)
- Distinct Class-specific Saliency Maps (DCSM): Distinct Class-specific Saliency Maps for Weakly Supervised Semantic Segmentation
- Weakly-Supervised Semantic Segmentation by Iteratively Mining Common Object Features
- Weakly Supervised Food Image Segmentation using Class Activation Maps
- Tell Me Where to Look: Guided Attention Inference Network
- Learning to Exploit the Prior Network Knowledge for Weakly-Supervised Semantic Segmentation
- Weakly-Supervised Semantic Segmentation Network with Deep Seeded Region Growing
- Seed, Expand and Constrain: Three Principles for Weakly-Supervised Image Segmentation (2016)
- Is object localization for free? – Weakly-supervised learning with convolutional neural networks (2015)
- LeNet-5 (1998): Gradient-Based Learning Applied to Document Recognition
- AlexNet (introduced dropout) (2012): ImageNet Classification with Deep Convolutional Neural Networks
- Network In Network (introduced Global Average Pooling) (2014): Network In Network
- Inception-V1/GoogLeNet (introduced convolutional blocks with smaller filters) (2014): Going deeper with convolutions
- VGG-16 and -19 (2015): Very Deep Convolutional Networks for Large-Scale Image Recognition
- ResNet (introduced skip-connections)(2015): Deep Residual Learning for Image Recognition
- Inception-V2 and -V3 (2015): Rethinking the Inception Architecture for Computer Vision
- Inception-V4 and -ResNet (2016): Inception-v4, Inception-ResNet and the Impact of Residual Connections on Learning
- Batch Normalization (2015): Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
- An Analysis of Deep Neural Network Models for Practical Applications
- Understanding the Effective Receptive Field in Deep Convolutional Neural Networks
- Receptive Fields Neural Networks using the Gabor Kernel Family
- Structured Receptive Fields in CNNs
- Invariant Scattering Convolution Networks
- Fully Convolutional Networks (FCN) - encoder-decoder architecture (2014): Fully Convolutional Networks for Semantic Segmentation
- U-Net - dilated/atrous convolution (2015): U-Net: Convolutional Networks for Biomedical Image Segmentation
- Fully connected CRF (2012): Efficient Inference in Fully Connected CRFs with Gaussian Edge Potentials
- SegNet (2015): SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation
- Dilated Convolutions (2016): Multi-Scale Context Aggregation by Dilated Convolutions
- DeepLab v1 (2014): Semantic Image Segmentation with Deep Convolutional Nets and Fully Connected CRFs
- DeepLab v2 (2016): DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs
- RefineNet (2016): RefineNet: Multi-Path Refinement Networks for High-Resolution Semantic Segmentation
- PSPNet (2016): Pyramid Scene Parsing Network
- Large Kernel Matters (2017): Large Kernel Matters -- Improve Semantic Segmentation by Global Convolutional Network
- DeepLab v3 (2017): Rethinking Atrous Convolution for Semantic Image Segmentation
- CRF-as-RNN (2015), consisting of a FCN followed by CRF: Conditional Random Fields as Recurrent Neural Networks
- G-CRF (2016), consisting of some segmenting network, followed by QO network: Fast, Exact and Multi-Scale Inference for Semantic Image Segmentation with Deep Gaussian CRFs
Source: A 2017 Guide to Semantic Segmentation with Deep Learning
- Guided Attention Inference Network (GAIN & GAIN_ext) (2018): Tell Me Where to Look: Guided Attention Inference Network