Autocoders are a family of neural network models aiming to learn compressed latent variables of high-dimensional data. Starting from the basic autocoder mode...
```iii) However, if you were to use your same Gaussian decoder to model data that is itself Gaussian, you'd find that the VAE learns to ignore the latent code!```
D. Wang, P. Cui, and W. Zhu. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD '16, page 1225--1234. San Francisco, California, USA, ACM Press, (2016)
W. Yu, C. Zheng, W. Cheng, C. Aggarwal, D. Song, B. Zong, H. Chen, and W. Wang. Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining - KDD '18, page 2663--2671. London, United Kingdom, ACM Press, (2018)
D. Wang, P. Cui, and W. Zhu. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD '16, page 1225--1234. San Francisco, California, USA, ACM Press, (2016)
D. Jin, B. Li, P. Jiao, D. He, and W. Zhang. Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence, page 2663--2669. Macao, China, International Joint Conferences on Artificial Intelligence Organization, (August 2019)
M. Rudolph, B. Wandt, and B. Rosenhahn. Third International Workshop on “Robust Subspace Learning and Applications in Computer Vision” (ICCV), (August 2019)