Build a Variational AutoEncoder (VAE) using PyTorch - Example using USPS dataset

In this video, I built a Variational AutoEncoder (VAE) from scratch using PyTorch. I used USPS dataset for this purpose which consists of digit images of very low resolution (16 x 16 spatial size).

01:28 - 𝓘𝓶𝓹𝓸𝓻𝓽 𝓽𝓱𝓮 𝓵𝓲𝓫𝓻𝓪𝓻𝓲𝓮𝓼
03:16 - 𝓓𝓮𝓯𝓲𝓷𝓮 𝓽𝓱𝓮 𝓭𝓮𝓿𝓲𝓬𝓮
05:29 - 𝓛𝓸𝓪𝓭 𝓤𝓢𝓟𝓢 𝓭𝓪𝓽𝓪𝓼𝓮𝓽
17:44 - 𝓓𝓮𝓯𝓲𝓷𝓮 𝓭𝓪𝓽𝓪𝓵𝓸𝓪𝓭𝓮𝓻𝓼
09:05 - 𝓓𝓮𝓯𝓲𝓷𝓮 𝓽𝓱𝓮 𝓷𝓮𝓽𝔀𝓸𝓻𝓴 𝓪𝓻𝓬𝓱𝓲𝓽𝓮𝓬𝓽𝓾𝓻𝓮
17:03 - 𝓥𝓲𝓼𝓾𝓪𝓵𝓲𝔃𝓮 𝓽𝓱𝓮 𝓶𝓸𝓭𝓮𝓵 𝓪𝓻𝓬𝓱𝓲𝓽𝓮𝓬𝓽𝓾𝓻𝓮
19:25 - 𝓣𝓻𝓪𝓲𝓷 𝓫𝓪𝓽𝓬𝓱
21:22 - 𝓣𝓮𝓼𝓽 𝓫𝓪𝓽𝓬𝓱
23:06 - 𝓢𝓮𝓽 𝓾𝓹 𝓶𝓸𝓭𝓮𝓵 (𝓥𝓐𝓔), 𝓬𝓻𝓲𝓽𝓮𝓻𝓲𝓸𝓷 (𝓵𝓸𝓼𝓼 𝓯𝓾𝓷𝓬𝓽𝓲𝓸𝓷) 𝓪𝓷𝓭 𝓸𝓹𝓽𝓲𝓶𝓲𝔃𝓮𝓻
26:07 - 𝓣𝓻𝓪𝓲𝓷 𝓽𝓱𝓮 𝓶𝓸𝓭𝓮𝓵
32:38 - 𝓟𝓵𝓸𝓽 𝓶𝓸𝓭𝓮𝓵 𝓵𝓸𝓼𝓼
34:36 - 𝓜𝓸𝓭𝓮𝓵 𝓹𝓻𝓮𝓭𝓲𝓬𝓽𝓲𝓸𝓷
40:05 - 𝓒𝓻𝓮𝓪𝓽𝓮 𝓼𝓪𝓶𝓹𝓵𝓮𝓼 𝓯𝓻𝓸𝓶 𝓷𝓸𝓲𝓼𝓮 𝓿𝓮𝓬𝓽𝓸𝓻𝓼

#data_science #deep_learning #jupyter_notebook #pytorch #variational_autoencoder #unsupervised_learning