ToDo papers

[Yagami360]

Basic models

• Neural Networks

  • [x] ODENet [Neural Ordinary Differential Equations]
  • [x] Augmented Neural ODEs
  • [ ] Kervolutional Neural Networks

• CNN

  • [x] Convolutional neural network architecture for geometric matching
  • [x] Spatial transformer networks

• Graph NN

  • [x] Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering
  • [x] Semi-Supervised Classification with Graph Convolutional Networks
  • [x] R-GCN [Relational Graph Convolutional Network]
  • [ ] Semi-supervised Learning with Graph Learning-Convolutional Networks

• GANs

  • [x] WGAN-gp
  • [x] InfoGAN
  • [ ] BigGAN
  • [x] Neural Collage [Spatially Controllable Image Synthesis with Internal Representation Collaging]
  • [x] [RGAN, RaGAN] The relativistic discriminator: a key element missing from standard GAN

• VAE

  • [x] VQ-VAE
  • [ ] VQ-VAE-2
  • [x] β-VAE

• normalization

  • [ ] Spectral normalization for generative adversarial networks

• Autoregressive Models

  • [x] PixelCNN, PixelRNN

• meta-learning, few-shot learning

  • [x] MAML:Model Agnostic Meta-Learning for Fast Adaption of Deep Networks

• Neural Processes

  • [ ] Deep Kernel Learning
  • [ ] Deep Neural Networks as Gaussian Processes
  • [x] Conditional Neural Processes
  • [x] Neural Processes
  • [ ] Attentive Neural Processes

Applications (CV)

• image classification

  • [ ] Squeeze-and-Excitation Networks

• semantic segmentation

  • [ ] Seg-Net
  • [ ] DeepLab v3
  • [ ] YOLACT: real-time instance segmentation
  • [ ] The Lovasz-Softmax loss
  • [x] Boundary loss for highly unbalanced segmentation
  • [ ] A Novel Focal Tversky loss function with improved Attention U-Net for lesion segmentation
  • human parsing
    • [ ] 候補（実装付き）: https://paperswithcode.com/task/human-parsing
    • [x] [JPPNet] Look into Person: Joint Body Parsing & Pose Estimation Network and A New Benchmark

• object detection

  • [ ] Feature pyramid networks for object detection
  • [x] Fast R-CNN
  • [x] Faster R-CNN: To- wards real-time object detection with region proposal networks
  • [ ] SSD
  • [ ] YOLO

• instance segmentation

  • [x] Mask r-cnn

• Pose Estimation

  • [x] DensePose: Dense Human Pose Estimation in the Wild

• Image Registration / geometric matching

  • [x] Spatial Transformer Networks
  • [x] [Geometric-matching CNN] Convolutional neural network architecture for geometric matching
  • [ ] End-to-End Unsupervised Deformable Image Registration with a Convolutional Neural Network
  • [ ] Quicksilver: Fast Predictive Image Registration – a Deep Learning Approach
  • [ ] VoxelMorph: A Learning Framework for Deformable Medical Image Registration

• image-to-image

  • [x] SinGAN: Learning a Generative Model from a Single Natural Image
  • [x] Pose Guided Person Image Generation

• inpainting

  • [x] Boundless: Generative Adversarial Networks for Image Extension
  • [x] Deepfill2 : Free-form image inpainting with gated convolution.
  • [ ] [Deepfillv1] Generative Image Inpainting with Contextual Attention

• virtual-try-on (2D base)

  • [ ] 候補（実装付き）:https://paperswithcode.com/task/virtual-try-on/latest#code
  • [ ] 候補 : https://scholar.google.co.jp/scholar?as_ylo=2019&hl=ja&as_sdt=2005&sciodt=0,5&cites=4604722267799491636&scipsc=
  • [ ] 候補 : https://scholar.google.co.jp/scholar?as_ylo=2019&hl=ja&as_sdt=2005&sciodt=0,5&cites=8505438031219592659&scipsc=
  • [x] VITON: An Image-based Virtual Try-on Network
  • [x] Virtually Trying on New Clothing with Arbitrary Poses
  • [x] Soft-gated warping-GAN for pose-guided person image synthesis
  • [x] FW-GAN: Flow-navigated Warping GAN for Video Virtual Try-on
  • [x] Robust Cloth Warping via Multi-Scale Patch Adversarial Loss for Virtual Try-On Framework
  • [x] ClothFlow: A Flow-Based Model for Clothed Person Generation
  • [ ] Be your own prada: Fashion synthesis with structural coherence.

• オプティカルフロー(Optical Flow)

  • [x] FlowNet: Learning Optical Flow with Convolutional Networks
  • [ ] Flownet 2.0: Evolution of optical flow estimation with deep networks
  • [ ] Optical flow estimation using a spatial pyramid network
  • [x] View synthesis by appearance flow

• 3D shape prediction

  • [ ] Shapenet: An information-rich 3d model repository
  • [ ] 3D-R2N2: A unified approach for single and multi-view 3d object reconstruction
  • [ ] A point set generation network for 3d object reconstruction from a single image
  • [ ] Pixel2Mesh: Generating 3D mesh models from single RGB images

• image-to-3D / image-based 3D Reconstruction

  • [ ] 候補（実装付き） : https://paperswithcode.com/task/3d-reconstruction
  • [ ] survey paper : Image-based 3D Object Reconstruction: State-of-the-Art and Trends in the Deep Learning Era
  • [ ] Learning category-specific mesh reconstruction from image collections
  • [ ] GEOMetrics: Exploiting geometric structure for graph-encoded objects
  • [ ] Pix3D : Learning to Reconstruct Shapes from Unseen Classes

• 3D-to-3D / 3D human generation

  • [x] SMPL: A Skinned Multi-Person Linear Model
  • [x] Multi-garment net: Learning to dress 3d people from images
  • [ ] 360-degree textures of people in clothing from a single image

• virtual-try-on (3D base)

  • [ ] Detailed Garment Recovery from a Single-View Image
  • [ ] ClothCap: Seamless 4D Clothing Capture and Retargeting

• texture mapping

  • [ ] Densereg: Fully convolutional dense shape regression in-the-wild

Theory

• [ ] Manifold Modeling in Embedded Space: A Perspective for Interpreting "Deep Image Prior"