Profile cover photo
Profile photo
Tobias Weyand
170 followers
170 followers
About
Tobias's posts

Post has shared content

Post has shared content
New LARGE dataset for object detection and tracking in videos!!
I hope this solves a long-standing gap in the space of computer vision datasets: object detection is hard, and most practical applications of it are in the context of video, not still images.
Videos enable you to reason about scene layout, persistence and occlusions, which should all really help in that context. Yet until now, most datasets either focussed on frame classification or whole-scene recognition.
After a LOT of work from my colleagues, here is finally a video dataset large enough and rich enough that you can train object detectors on it. Bonus: the videos are intentionally selected for their 'naturalness', which means they should transfer to real-world use cases with minimal domain shift. Enjoy!

Post has shared content
I'm excited to share this post about one of my projects at Google!
Think this is a picture of the Eiffel Tower in Paris? The Vision API isn't fooled. Learn more about how landmark detection works: goo.gl/sR4Ya7
Photo

Post has shared content

Post has shared content

Post has attachment
I'm excited that I can finally share what I've been working on at Google! We just published our paper describing PlaNet, a Convolutional Neutral Network that can tell where in the world a photo was taken. From millions of geotagged photos PlaNet has learned which countries have fjords, where to find red phone booths, and which animals live on the Galapagos Islands. Check out our paper on arXiv below! A big thank you goes to my co-authors +Ilya Kostrikov and +James Philbin!

Post has shared content
Google's landmark recognition (and more!) is now available to everyone!
Now you can detect objects in your images the same way +Google Photos does. Learn how to get started with Cloud Vision API: http://goo.gl/iePJsA
Photo

Post has shared content
Turning black-and-white photographs into colour photographs with a bit of deep learning.
I really enjoyed this one - it is cool and nicely written. The author is the original designer of node.js. As far as I know he doesn't have a background in machine learning. I wonder if magic will soon be in every good programmer's toolbox?

Post has shared content
The next generation of applications won’t just store images. They’ll actually understand them thanks to models like the one we’re releasing today. Our Cloud Vision API uses machine learning to classify images into thousands of categories, detect objects (and facial features), and even identify language. We’re excited to see what you build with it: http://goo.gl/RfEFv0
Photo

Post has shared content
Wait while more posts are being loaded