課程簡介
GAN生成式對抗網(wǎng)絡(luò)
目標(biāo)收益
培訓(xùn)對象
對深度學(xué)習(xí)算法原理和應(yīng)用感興趣，具有一定編程(Python)和數(shù)學(xué)基礎(chǔ)(線性代數(shù)、微積分、概率論)的技術(shù)人員。
對深度學(xué)習(xí)模型，特別是生成式模型（有一定了解為佳。
課程大綱
| 1. GAN 入門 - Generative Models |
- Latent Factors - Generative Models |
| 2. GAN 原理 |
- Discrimination and Generator - Training GAN - Distances in GAN: from KL-Divergence to JS-Divergence and others - Problems with GAN - CGAN and DCGAN |
| 3. f-GAN 模型 - GAN 模型的同一框架 |
- Fenchel Conjugate - f-Divergence - Training: double loop vs single loop - 多種 divergence 函數(shù) |
| 4. Wasserstein GAN - WGAN |
- Problem with JS-Divergence - Mode Collapse - Earth-Mover Distance - WGAN - EBGAN: Energy-Based GAN |
| 5. InfoGAN - 可解釋表示的 GAN |
- 潛因子與表象的互信息 - 現(xiàn)有 GAN 和 Domain 之間的矛盾 - 用無監(jiān)督學(xué)習(xí)發(fā)現(xiàn)可解釋的潛因子 - 帶互信息正則項的 loss 函數(shù) - 實現(xiàn)：用變分法進(jìn)行訓(xùn)練 - 效果 |
| 6. GAN for NLP |
- Improving Sequence Generation with GAN - Conditional Sequence Generation - Unsupervised Conditional Sequence Generation |
| 7. GAN for CV |
- GAN + Autoencoder: Photo Editing - Image Super Resolution - Image Completion |
|
1. GAN 入門 - Generative Models - Latent Factors - Generative Models |
|
2. GAN 原理 - Discrimination and Generator - Training GAN - Distances in GAN: from KL-Divergence to JS-Divergence and others - Problems with GAN - CGAN and DCGAN |
|
3. f-GAN 模型 - GAN 模型的同一框架 - Fenchel Conjugate - f-Divergence - Training: double loop vs single loop - 多種 divergence 函數(shù) |
|
4. Wasserstein GAN - WGAN - Problem with JS-Divergence - Mode Collapse - Earth-Mover Distance - WGAN - EBGAN: Energy-Based GAN |
|
5. InfoGAN - 可解釋表示的 GAN - 潛因子與表象的互信息 - 現(xiàn)有 GAN 和 Domain 之間的矛盾 - 用無監(jiān)督學(xué)習(xí)發(fā)現(xiàn)可解釋的潛因子 - 帶互信息正則項的 loss 函數(shù) - 實現(xiàn)：用變分法進(jìn)行訓(xùn)練 - 效果 |
|
6. GAN for NLP - Improving Sequence Generation with GAN - Conditional Sequence Generation - Unsupervised Conditional Sequence Generation |
|
7. GAN for CV - GAN + Autoencoder: Photo Editing - Image Super Resolution - Image Completion |
近期公開課推薦
