• Cloud TPU support: You can now use TF-GAN to train GANs on Google’s Cloud TPUs. Tensor Processing Units (TPUs) are Google’s custom-developed application-specific integrated circuits (ASICs) used to accelerate machine learning workloads. Models that previously took weeks to train on other hardware platforms can converge in hours on TPUs. This open source example demonstrates training an image-generation GAN on ImageNet on TPU, which we discuss in more detail below. You can also run a TF-GAN on TPU tutorial in colaboratory for free.
• Self-study GAN course: Machine learning works best when knowledge is freely available. To that end, we’ve released a self-study GAN course based on the GAN courses already taught internally at Google for years. Watching the videos, reading the descriptions, following the exercising, and doing the code examples are good steps on your road to ML mastery.
• GAN metrics: Academic papers sometimes “invent a ruler”, then use it to measure their results. To facilitate comparing results across papers, TF-GAN has made using standard metrics even easier. In addition to correcting for numerical precision and statistical biases that sometimes plague even the standard open-source implementations, TF-GAN metrics are computationally-efficient and syntactically easy to use.
• Examples: GAN research is incredibly fast-paced. While TF-GAN doesn’t intend to keep working examples of all GAN models, we have added some that we think are relevant, including a Self-Attention GAN that trains on TPU.
• PyPi package: TF-GAN can now be installed with ‘pip install tensorflow-gan’ and used with ‘import tensorflow_gan as tfgan’. Simple, right?
• Colaboratory tutorials: We’ve improved our tutorials. They can now be used with Google’s free GPUs and TPUs.
• GitHub Repository: TF-GAN is now in it’s own repository. This makes it easier to track changes and properly give credit to open-source contributors.
• TensorFlow 2.0: TF-GAN is currently TF 2.0 compatible, but we’re continuing to make it compatible with Keras. You can find some GAN Keras examples that don’t use TF-GAN at tensorflow.org, including DCGAN, Pix2Pix, and CycleGAN.

Projects using TF-GAN

Left: Real Middle: Generated (GPU) Right: Generated (TPU)

Left: Frechet Inception Distance and Inception score as a function of training step. Right: Frechet Inception Distance and Inception score as a function of training time.

Image Extension

Some examples of image extension: The new method, using TF-GAN, is in the right column. It generates better object shapes (top/middle rows) and produces good textures (middle/bottom rows), compared with two state of the art inpainting methods: DeepFill and PConv. The input image is extended onto the masked area (shown in gray, left column).

BigGAN

The DeepMind research team improved state-of-the-art image generation in this paper, using a combination of architectural changes, a larger network, larger batch sizes, and Google TPUs. They used TF-GAN’s evaluation module to standardize their metrics, and were able to demonstrate quality improvements across a wide range of image sizes. Pretrained BigGAN generators are available on TF Hub.

Class conditional examples generated from Large Scale GAN Training for High Fidelity Natural Image Synthesis.

Microscope image normalization with GANs

The image shows the output of the Equalizer applied to one cell slide image morphed into each of the 9 weeks of the experiment. Some slide properties are changed, such as average background color, but properties like ‘number of cells in the image’ remain the same.