Large Scale GAN Training for High Fidelity Natural Image Synthesis
At the seminar we will discuss recent paper "Large Scale GAN Training for High Fidelity Natural Image Synthesis".
Paper abstract: "Despite recent progress in generative image modeling, successfully generating high-resolution, diverse samples from complex datasets such as ImageNet remains an elusive goal. To this end, we train Generative Adversarial Networks at the largest scale yet attempted, and study the instabilities specific to such scale. We find that applying orthogonal regularization to the generator renders it amenable to a simple "truncation trick", allowing fine control over the trade-off between sample fidelity and variety by truncating the latent space. Our modifications lead to models which set the new state of the art in class-conditional image synthesis. When trained on ImageNet at 128x128 resolution, our models (BigGANs) achieve an Inception Score (IS) of 166.3 and Frechet Inception Distance (FID) of 9.6, improving over the previous best IS of 52.52 and FID of 18.65."
Speaker: Aleksei Shpilman.
Presentation language: Russian.
Date and time: February 13th, 18:30-20:00.
Location: Times, room 204.
Videos from previous seminars are available at http://bit.ly/MLJBSeminars
- About seminars
17 April 2019Open Questions about Generative Adversarial Networks
10 April 2019Adaptive Sampled Softmax with Kernel Based Sampling
3 April 2019Visual Object Tracking
13 March 2019Knowledge Distillation
6 March 2019BERT