In a few hours we are going to premier a video on JAX, Trax and why it's helpful to use these frameworks to train German TTS model - Speech Processing with Deep Learning and JAX/Trax - youtube.com/watch?v=00qwLpXX… by @GoogleDevExpert M. Yusuf Sarıgöz
computational photography makes so much progress these days. It's also pretty awesome, that @GoogleAI shares such articles on how such features are implemented
Today we present the technology behind the new Cinematic photos feature in Google Photos, where #MachineLearning is used to add simulated camera motion and parallax to a single-frame still photo, transforming a 2D photo into a more immersive 3D scene. goo.gle/3aKeRKr
Super excited to be hosting the next version of our event. One the talks would be on Speech processing with JAX and Trax using @nvidia's Tacotron2 and
MelGAN
New code walkthrough on keras.io: image search with natural language queries using a dual-encoder approach. The idea is to learn a joint embedding space for images and associated captions (inspired by OpenAI CLIP).
keras.io/examples/nlp/nl_ima…
Been wondering what is Contrastive Learning, or can't keep track of its development?
Excited to share my new (first 👶) paper
**Contrastive Representation Learning: A Framework and Review**, with Graham Healy and @asmeatonieeexplore.ieee.org/document…
or arxiv.org/abs/2010.05113 1/n
It's one type of hard when you have all the data and need to tune the SOTA. In real life there are way more other issues to face on your way to deploy ML to production. Super excited to host @kweinmeister and Anant Nawalgaria later tonight! #MachineLearning#tfug Link below 👇
After yesterday's talk of @haifeng_jin at #MLconfEU I looked at AutoKeras. And it looks like it has the friendliest automl api I've seen so far. I do not include simple examples, but more advanced with multi-modal, preprocessing and so on #tensorflow#keras#automl