|September 27 · Issue #8 · View online |
Wow, we really needed to whittle it down this week!
This industry is full of individual powerhouses innovating, researching, optimizing and sharing new technology, applications, and knowledge.
We were particularly impressed this week by neural photo editing (magic!) and the quality of teaching at last weekend’s Deep Learning School, the video recordings of both days are available on youtube.
Here’s to hoping next week will be half as exciting as this one.
As always if you enjoy receiving this newsletter, please considering sharing it with friends and colleagues, your support is very much appreciated.
| Google acquires natural language understanding startup Api.ai | VentureBeat | Bots | by Jordan Novet |
Google today said that it’s acquired API.ai, a startup with tools for speech recognition and natural language understanding. Terms of the deal weren’t disclosed.
| Deep Learning Obstacles: What's the Lesson? |
Broad overview of Google’s research endeavors in the area of Artificial Intelligence, with a specific focus on Deep Learning.
| Putting patients at the heart of DeepMind Health | DeepMind |
Thoughts and outcomes following DeepMind’s first open patient and public forum.
| Baidu Releases AI Benchmark | EE Times |
Calling for 100x faster processors, Baidu released DeepBench, a benchmark for how fast processors train neural networks.
| Object Detection using Deep Learning for advanced users (Part-1) |
Part 1 of a three-part series on the state of the art in Object Detection in images using Convolutional Neural Networks.
| GitHub - alrojo/tensorflow-tutorial: Practical tutorials and labs for TensorFlow used by Nvidia, FFN, CNN, RNN, Kaggle, AE |
tensorflow-tutorial - Practical tutorials and labs for TensorFlow used by Nvidia, FFN, CNN, RNN, Kaggle, AE
| Deep Learning School |
Two day event with an incredible line up of speakers including Yoshua Bengio, Hugo Larochelle, Andrew Ng and Andrej Karpathy, video lectures are available on youtube.
| Prof. Yann LeCun - Deep Learning and the Future of AI - YouTube |
Yann LeCun, Director of AI Research at Facebook on the future of AI.
| Show and Tell: image captioning open sourced in TensorFlow |
Google Research open sources the latest version of their image captioning system in TensorFlow
Improvements allow more detailed and accurate descriptions
| GitHub - tomlepaine/fast-wavenet: Efficient implementation of Wavenet generation. |
fast-wavenet - Efficient implementation of Wavenet generation.
| GitHub - openai/cleverhans: A library for benchmarking vulnerability to adversarial examples |
cleverhans - A Python library for benchmarking vulnerability to adversarial examples
| GitHub - junyanz/iGAN: A deep learning software that easily generates images with a few brushstrokes (from UC Berkeley and Adobe CTL) |
iGAN - A deep learning software that easily generates images with a few brushstrokes (from UC Berkeley and Adobe CTL)
| Show and Tell: Lessons learned from the 2015 MSCOCO Image Captioning Challenge |
Presentation of the generative model based on a deep recurrent architecture used by Google for their image captioning library.
| Neural Photo Editing with Introspective Adversarial Networks |
Fascinating approach using an Introspective Adversarial Network to make large, semantically coherent changes to existing images. Watch the demo here: https://www.youtube.com/watch?v=FDELBFSeqQs
| Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation |
Google introduces GNMT (Google’s Neural Machine Translation system) consisting of a deep LSTM network designed to overcome many of the current challenge in neural machine translation.