|March 21 · Issue #77 · View online |
Welcome back to another tempestuous week in Deep Learning.
Happy reading and hacking!
| Inside the Chinese Lab that Plans to Rewire the World with AI |
Alibaba is investing huge sums in AI research and resources—and it is building tools to challenge Google and Amazon. Alibaba’s ambition is to be the leader in providing cloud-based AI. Like cloud storage (think Dropbox) or cloud computing (Amazon Web Services), cloud AI will make powerful resources cheaply and readily available to anyone with a computer and an internet connection, enabling new kinds of businesses to grow.
| Ingredients for Robotics Research |
OpenAI releases eight simulated robotics environments and a baselines implementation of Hindsight Experience Replay, all developed for their research over the past year. These environments can be used to train models which work on physical robots.
| Deep Cognition is Launching Deep Learning Studio 2.0 at NVIDIA’s GTC 2018 (sponsored) |
Open, Free and No-coding deep learning platform is getting an upgrade. Check out what is coming next ( Hint: If are a developer then you will love us more).
| Transfer Your Font Style with GANs |
Berkeley researchers present GlyphNet a the multi-content GAN architecture which retrains a customized magical network for each observed character set with only a handful of observed glyphs. This model considers content (i.e., A-Z glyphs) along channels and style (i.e., glyph ornamentations) along network layers to transfer the style of given glyphs to the contents of unseen ones.
| The Machine Learning Reproducibility Crisis |
A fascinating an unsettling take on the machine learning reproducibility crisis by Pete Warden who outlines the cycle of machine learning models which makes them so hard to track changes or collaborate on.
| Compiling Deep Learning Models to WebGL with TVM |
TVM already targets a lot of backends covering a variety of platforms: CPU, GPU, mobile devices, etc. OpenGL/WebGL enables us to leverage the GPU on an environment which does not have CUDA installed. It is, for the time being, the only way of using the GPU inside a browser.
| Diving into Deep Convolutional Semantic Segmentation Networks and Deeplab_V3 |
This piece provides an introduction to Semantic Segmentation with a hands-on TensorFlow implementation, it goes over one of the most relevant papers on Semantic Segmentation of general objects — Deeplab_v3
| Making Music Using New Sounds Generated with Machine Learning |
Interesting post introducing Magenta, a research project within Google which released NSynth (Neural Synthesizer) last year, a machine learning algorithm that uses deep neural networks to learn the characteristics of sounds, and then create a completely new sound based on these characteristics. As part of this exploration, Google Creative Lab and Magenta collaborated to create NSynth Super, an open source experimental instrument which gives musicians the ability to explore new sounds generated with the NSynth algorithm.
| MusicVAE: Creating a Palette for Musical Scores with Machine Learning. |
Another piece of fascinating work by the Google Magenta team; MusicVAE is a machine learning model that lets artists and composers create palettes for blending and exploring musical scores. MusicVAE is based on latent space models a class of models to represent the variation in a high-dimensional dataset using a lower-dimensional code, making it easier to explore and manipulate intuitive characteristics of the data.
| Baidu Apollo Releases Massive Self-driving Dataset; Teams Up With Berkeley DeepDrive |
Baidu this Thursday announced the release of Apollo Scape, billed as the world’s largest open-source dataset for autonomous driving technology. The data volume of ApolloScape is 10 times greater than any other open-source autonomous driving dataset, including Kitti and CityScapes.
| labNotebook: an Experiment Manager for Deep Learning Experiments |
labNotebook is a simple experiment manager for deep learning experiments, it allows you to flexibly save all your experimental data in a Postgres database through a simple interface, including configuration, models, results, and training curves and to monitor any indicators from your running experiments by streaming them through a web application.
| VINE: An Open Source Interactive Data Visualization Tool for Neuroevolution |
Uber AI Labs introduces Visual Inspector for Neuroevolution (VINE), an open source interactive data visualization tool to help neuroevolution researchers better understand this family of algorithms.
| Data Scientist |
Deep NLP and Vision for Business at Infinia ML.
| Data Scientist Intern |
Deep Learning for Business at Infinia ML.
| Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge |
The authors present a new question set, text corpus, and baselines assembled to encourage AI research in advanced question answering. Together, these constitute the AI2 Reasoning Challenge (ARC), which requires far more powerful knowledge and reasoning than previous challenges such as SQuAD or SNLI.
| Microsoft AI research: Achieving Human Parity on Automatic Chinese to English News Translation |
Microsoft AI research reports that they’ve achieved human parity on Chinese-to-English translation. The quality of Microsoft’s machine translation system translations is measured on the widely used WMT 2017 news translation task from Chinese to English. The authors find that the system has reached a new state-of-the-art, and that the translation quality is at human parity when compared to professional human translations.