Distributed Representations of Words and Phrases and their Compositionality

Several extensions to #SkipGram that improve both the quality of the vectors and the training speed

ArXiV: https://arxiv.org/abs/1310.4546

#NLP #oldenough

​​Real-Time Patch-Based Stylization of Portraits Using Generative Adversarial Network

Face photo stylization from #Snap research team. Rather fast solution with demo available.

Demo: http://facestyle.org/#
Paper: https://dcgi.fel.cvut.cz/home/sykorad/Futschik19-NPAR.pdf
YouTube: https://www.youtube.com/watch?v=G3nwTSd3_XA

#GAN #DL #Styletransfer

Speech synthesis from neural decoding of spoken sentences

Researchers tapped the brains of five epilepsy patients who had been implanted with electrodes to map the source of seizures, according to a paper published by #Nature. During a lull in the procedure, they had the patients read English-language texts aloud. They recorded the fluctuating voltage as the brain controlled the muscles involved in speaking. Later, they fed the voltage measurements into a synthesizer.

Nature: https://www.nature.com/articles/s41586-019-1119-1
Paper: https://www.biorxiv.org/content/biorxiv/early/2018/11/29/481267.full.pdf
YouTube: https://www.youtube.com/watch?v=kbX9FLJ6WKw

#DeepDiveWeekly #DL #speech #audiolearning

Unsupervised community detection with modularity-based attention model

Searching for communities on graphs is hard -> no clear loss, discrete labels (usually). What we do: use soft log-liklehood approximation with tricks + GNNs to try to match classical SOTA.

Paper: https://rlgm.github.io/papers/37.pdf

#ICLR2019 #GNN #GraphLearning

The lottery ticket hypothesis: finding sparse, trainable neural networks

Best paper award at #ICLR2019 main idea: dense, randomly-initialized, networks contain sparse subnetworks that trained in isolation reach test accuracy comparable to the original network. Thus compressing the original network up to 10% its original size.

Paper: https://arxiv.org/pdf/1803.03635.pdf

#nn #research

​​Modeling Price with Regularized Linear Model & #XGBoost

Great example of applicable research for #production #ML.

Link: https://www.kdnuggets.com/2019/05/modeling-price-regularized-linear-model-xgboost.html

#novice #entrylevel

The Feynman Lectures on Physics now free (Score: 100+ in 17 hours)

Link: https://readhacker.news/s/43fFn
Comments: https://readhacker.news/c/43fFn

​​🐣 Conversational AI building tutorial, open-source code & demo!

Building a SOTA Conversational AI with transfer learning & OpenAI GPT models

Code/pretrained model from NeurIPS 2018 ConvAI2 competition model, SOTA on automatic track
Detailed Tutorial w. code

Tutorial: https://medium.com/huggingface/how-to-build-a-state-of-the-art-conversational-ai-with-transfer-learning-2d818ac26313
Demo: http://convai.huggingface.co
Code: https://github.com/huggingface/transfer-learning-conv-ai

#tutorial #novice

All postres from #ICLR2019 gathered in one place

Link: https://postersession.ai

​​End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography

Researchers from #GoogleAi and #Stanford published work today in #Nature that shows great potential to use machine learning to help catch more lung cancer cases earlier and increase survival likelihood.

Link: http://go.nature.com/2LSMaAz

#LungCancer #Cancer #biolearning #healthcare #DL #CV

New deep learning framework from Facebook

Pythia is a deep learning framework that supports multitasking in the vision and language domain. Built on our open-source #PyTorch framework, the modular, plug-and-play design enables researchers to quickly build, reproduce, and benchmark AI models. #Pythia is designed for vision and language tasks, such as answering questions related to visual data and automatically generating image captions.

Link: https://code.fb.com/ai-research/pythia/
GitHub: https://github.com/facebookresearch/pythia

#Facebook #FacebookAI #DL #CV #multimodal

Parisian Open Data Science branch is growing!

Come to have a coffee with a fellow Data Scientists. We speak English, Russian or French. This Saturday, 25th of May, at 10h30 at Malongo Cafe, 50 Rue Saint-André des Arts.

Parallel Neural Text-to-Speech

#Baidu Research team has reached a new milestone in text-to-speech (#TTS) with the release of the first fully parallel neural TTS system. It generates speech from text with a single feed-forward pass, which brings ~17.5X speed-up over previous autoregressive models.

Link: https://arxiv.org/pdf/1905.08459.pdf

​​Head animation from single shot by #SamsungAI team

Samsung researchers have released a model that can generate faces in new poses from just a single image/frame (for each of face, pose). Done by building a well-trained landmark model in advance & one-shotting from that, using keypoints, adaptive instance norms and GANs. Model performs no 3D face modelling!

ArXiV: https://arxiv.org/abs/1905.08233v1
Youtube: https://www.youtube.com/watch?v=p1b5aiTrGzY

#GAN #CV #DL

​​Recovering person appereance from person’s speech

As the result of the research, much resembling facial image of a person reconstructed from short audio recording of that person speaking.

ArXiV: https://arxiv.org/pdf/1905.09773v1.pdf

#speech #audiolearning #CV #DL #face

🇮🇹If by any chance you are in Venice for week or two and would like to meet and talk about Data Science and have some caffe normale, pm @malev.

​​Set of animated Artificial Intelligence cheatsheets covering Stanford's CS 221 class:

Reflex-based: http://stanford.io/2EqNPHy 
States-based: http://stanford.io/2wh4F7u 
Variables-based: http://stanford.io/2HAiAfh 
Logic-based: http://stanford.io/2M7taia 

GitHub: https://github.com/afshinea/stanford-cs-221-artificial-intelligence

#cheatsheet #Stanford

MetroTwitter - What Twitter reveals about the differences between cities and the monoculture of the Bay Area

Researcher collected 96K bios + 180M tweets from Twitters users in 13 major cities and visualized the differences between these cities:

- How people describe themselves
- What they talk about
- Popular emojis
- Most unique city

Code and data are open-sourced.

Website: https://huyenchip.com/2019/05/28/metrotwitter.html
GitHub: https://github.com/chiphuyen/MetroTwitter

#openresearch

Small $ 60 camera for #CV tasks

JeVois — video sensor + quad-core CPU + USB video + serial port, all in a tiny, self-contained package size of a small coin.


YouTube: https://www.youtube.com/watch?v=7cTfOckkGlE
Project site: http://jevois.org
Link for shop: https://www.jevoisinc.com


#deeplearning #hardware #tensorflow

Panel: A high-level app and dashboarding solution for the PyData ecosystem.

New framework for handling python dashboards.

Link: https://medium.com/@philipp.jfr/panel-announcement-2107c2b15f52

#vizualization #panel #metrics #dashboard #python