個人的に面白かったMachine Learning論文 in 2019 Part 3 — GAN,実応用 —

  • Part 1 : 画像・動画系と学習の工夫関連
  • Part 2 : NLP、自然科学分野、DLの解析
  • Part 3 : GAN、実社会応用、その他分野(この記事)

1. GAN(Style Transfer関連)

  • SEAN: Image Synthesis with Semantic Region-Adaptive Normalization
  • TransGaGa: Geometry-Aware Unsupervised Image-to-Image Translation
  • Analyzing and Improving the Image Quality of StyleGAN
  • SC-FEGAN Face Editing Generative Adversarial Network with User’s Sketch and Color
  • MarioNETte: Few-shot Face Reenactment Preserving Identity of Unseen Targets
  • FaceShifter: Towards High Fidelity And Occlusion Aware Face Swapping
  • Few-shot Video-to-Video Synthesis
  • Few-Shot Adversarial Learning of Realistic Neural Talking Head Models

2. GAN(生成画像の質改善/新アーキテクチャ/新タスク設定)

  • High-Fidelity Image Generation With Fewer Labels
  • Consistency Regularization for Generative Adversarial Networks
  • SMALL-GAN: SPEEDING UP GAN TRAINING USING CORE-SET
  • LOGAN: LATENT OPTIMISATION FOR GENERATIVE ADVERSARIAL NETWORKS
  • HoloGAN: Unsupervised learning of 3D representations from natural images
  • SinGAN: Learning a Generative Model from a Single Natural Image
  • Interactive Sketch & Fill: Multiclass Sketch-to-Image Translation

3. 実社会の応用

  • Counterfactual Visual Explanations
  • GRAPE DETECTION, SEGMENTATION AND TRACKING USING DEEP NEURAL NETWORKS AND THREE-DIMENSIONAL ASSOCIATION
  • CITY METRO NETWORK EXPANSION WITH REINFORCEMENT LEARNING

4. その他

  • What’s Hidden in a Randomly Weighted Neural Network?
  • Making the Invisible Visible: Action Recognition Through Walls and Occlusions
  • GraphNVP: An Invertible Flow Model for Generating Molecular Graphs
  • How to make a pizza- Learning a compositional layer-based GAN model
  • Superposition of many models into one
  • Go-Explore: a New Approach for Hard-Exploration Problems
  • The Domain Shift Problem of Medical Image Segmentation and Vendor-Adaptation by Unet-GAN

1. GAN(Style Transfer関連)

1.1. SEAN: Image Synthesis with Semantic Region-Adaptive Normalization

1.2. TransGaGa: Geometry-Aware Unsupervised Image-to-Image Translation

1.3. Analyzing and Improving the Image Quality of StyleGAN

1.4. SC-FEGAN Face Editing Generative Adversarial Network with User’s Sketch and Color

1.5. MarioNETte: Few-shot Face Reenactment Preserving Identity of Unseen Targets

1.6. FaceShifter: Towards High Fidelity And Occlusion Aware Face Swapping

1.7. Few-shot Video-to-Video Synthesis

1.8. Few-Shot Adversarial Learning of Realistic Neural Talking Head Models

2. GAN(生成画像の質改善/新アーキテクチャ/新タスク設定)

2.1. High-Fidelity Image Generation With Fewer Labels

2.2. Consistency Regularization for Generative Adversarial Networks

2.3. SMALL-GAN: SPEEDING UP GAN TRAINING USING CORE-SET

2.4. LOGAN: LATENT OPTIMISATION FOR GENERATIVE ADVERSARIAL NETWORKS

2.5. HoloGAN: Unsupervised learning of 3D representations from natural images

2.6. SinGAN: Learning a Generative Model from a Single Natural Image

2.7. Interactive Sketch & Fill: Multiclass Sketch-to-Image Translation

3. 実社会への応用

3.1. Counterfactual Visual Explanations

3.2. GRAPE DETECTION, SEGMENTATION AND TRACKING USING DEEP NEURAL NETWORKS AND THREE-DIMENSIONAL ASSOCIATION

3.3. CITY METRO NETWORK EXPANSION WITH REINFORCEMENT LEARNING

4. その他

4.1. What’s Hidden in a Randomly Weighted Neural Network?

4.2. Making the Invisible Visible: Action Recognition Through Walls and Occlusions

4.3. GraphNVP: An Invertible Flow Model for Generating Molecular Graphs

4.4. How to make a pizza- Learning a compositional layer-based GAN model

4.5. Superposition of many models into one

4.6. Go-Explore: a New Approach for Hard-Exploration Problems

4.7. The Domain Shift Problem of Medical Image Segmentation and Vendor-Adaptation by Unet-GAN

まとめ

--

--

--

Data Scientist (Engineer) in Japan Twitter : https://twitter.com/AkiraTOSEI LinkedIn : https://www.linkedin.com/mwlite/in/亮宏-藤井-999868122

Love podcasts or audiobooks? Learn on the go with our new app.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Akihiro FUJII

Akihiro FUJII

Data Scientist (Engineer) in Japan Twitter : https://twitter.com/AkiraTOSEI LinkedIn : https://www.linkedin.com/mwlite/in/亮宏-藤井-999868122

More from Medium

The Relevance Between Themes In Of Mice And Men And Todays Society

Overwhelm

Is conflict inevitable?