SchoolOfMa :: 7/19/2016
Applications of convnets
- Review how convnets work (0:00)
- Visualizing and interpreting learned features (21:54)
- Why the activations are valuable (27:46)
- Image embeddings with t-SNE (36:38)
- Class visualization and Deepdream (53:37)
- Style transfer (1:25:12)
- Special cases of style transfer: video, real-time, colorless (1:42:14)
- Image analogies, neural-doodle, super-resolution, assistive (1:49:10)
- Colorizing black & white images (1:55:27)
- Deep convolutional generative adversarial networks (1:57:34)
- Tutorial: neural-style in terminal instance (2:03:01)