Preatty good lecture about recommendation systems https://www.youtube.com/watch?v=P1tJuzthqPA
YouTube
Михаил Камалов — Рекомендательные системы: от матричных разложений к глубинному обучению
—
В настоящее время рекомендательные системы активно применяются как в сфере развлечений (YouTube, Netflix), так и в сфере интернет-маркетинга (Amazon, Aliexpress). В связи с этим, в докладе будут рассмотрены практические аспекты применения глубинного обучения…
В настоящее время рекомендательные системы активно применяются как в сфере развлечений (YouTube, Netflix), так и в сфере интернет-маркетинга (Amazon, Aliexpress). В связи с этим, в докладе будут рассмотрены практические аспекты применения глубинного обучения…
New cool lib for dimension reduction. Better then largeVis
https://github.com/lmcinnes/umap
https://github.com/lmcinnes/umap
GitHub
GitHub - lmcinnes/umap: Uniform Manifold Approximation and Projection
Uniform Manifold Approximation and Projection. Contribute to lmcinnes/umap development by creating an account on GitHub.
Libs that will help if you are working with sentence embedding
https://github.com/facebookresearch/InferSent
https://github.com/facebookresearch/SentEval
https://github.com/facebookresearch/InferSent
https://github.com/facebookresearch/SentEval
GitHub
GitHub - facebookresearch/InferSent: InferSent sentence embeddings
InferSent sentence embeddings. Contribute to facebookresearch/InferSent development by creating an account on GitHub.
Updated SSD with higher accuracy in small object detection https://arxiv.org/abs/1711.07767
Very hot article about deep video fakes
https://web.stanford.edu/~zollhoef/papers/SG2018_DeepVideo/page.html
https://web.stanford.edu/~zollhoef/papers/SG2018_DeepVideo/page.html
This library will make you video loading faster as never before)
https://github.com/NVIDIA/nvvl
https://github.com/NVIDIA/nvvl
GitHub
GitHub - NVIDIA/nvvl: A library that uses hardware acceleration to load sequences of video frames to facilitate machine learning…
A library that uses hardware acceleration to load sequences of video frames to facilitate machine learning training - NVIDIA/nvvl
Interesting way of training GANS. Increasing image resolution with adding new layers.
https://github.com/tkarras/progressive_growing_of_gans
https://github.com/tkarras/progressive_growing_of_gans
GitHub
GitHub - tkarras/progressive_growing_of_gans: Progressive Growing of GANs for Improved Quality, Stability, and Variation
Progressive Growing of GANs for Improved Quality, Stability, and Variation - tkarras/progressive_growing_of_gans
Implementation of some papers from different ML areas
https://github.com/GauravBh1010tt/DeepLearn
https://github.com/GauravBh1010tt/DeepLearn
GitHub
GitHub - GauravBh1010tt/DeepLearn: Implementation of research papers on Deep Learning+ NLP+ CV in Python using Keras, Tensorflow…
Implementation of research papers on Deep Learning+ NLP+ CV in Python using Keras, Tensorflow and Scikit Learn. - GitHub - GauravBh1010tt/DeepLearn: Implementation of research papers on Deep Learni...
Strange to see this in Microsoft repo, but really usefull thing for crossframework model convertion https://github.com/Microsoft/MMdnn?utm_source=mybridge&utm_medium=blog&utm_campaign=read_more
GitHub
GitHub - microsoft/MMdnn: MMdnn is a set of tools to help users inter-operate among different deep learning frameworks. E.g. model…
MMdnn is a set of tools to help users inter-operate among different deep learning frameworks. E.g. model conversion and visualization. Convert models between Caffe, Keras, MXNet, Tensorflow, CNTK, ...
Прямо сейчас ребята рассказывают как заняли призовые места на Kaggle по Fraud detection and Nucleos detection
https://www.facebook.com/1616285355341854/videos/1839957856307935
https://www.facebook.com/1616285355341854/videos/1839957856307935
Facebook Watch
Rails Reactor
ML Training #4
New paper from Facebook Research team on human movement modeling with RL
https://github.com/facebookresearch/QuaterNet
https://github.com/facebookresearch/QuaterNet
GitHub
GitHub - facebookresearch/QuaterNet: Proposes neural networks that can generate animation of virtual characters for different actions.
Proposes neural networks that can generate animation of virtual characters for different actions. - facebookresearch/QuaterNet
Found interesting course about RL, with tons of practical examples https://simoninithomas.github.io/Deep_reinforcement_learning_Course/
Great news for those who likes Summer, Sea, and Data Science - Data Summer Conf 2018 starting on July 21.
2 streams: BigData and DataScience
Speakers from the world's biggest companies: AWS, Spotify, GridGain Systems, Google Developers
Giuseppe Angelo Porcelli, Solutions Architect at Amazon Web Services.
- Jonathan Taws, Data Scientist at Amazon Web Services;
- Sri Sri Perangur, Senior Data Scientists at Spotify
- Javier Rodriguez Zaurin, Head of Data Science at Simply Business;
- Akmal Chaudhri, Lead Technical Evangelist at GridGain Systems;
- Jacek Laskowski, Spark & Kafka Developer and Technical Instructor at Kafka Streams Consultant;
- Fedor Navruzov, Senior Data Scientist at SWAG Speak With A Geek;
- Giorgi Jvaridze, Software Engineer at Data Science Stream;
- Roman Strochak, Computer vision CTO at Data AI;
- Rudradeb Mitra, Product Mentor at Google Developers;
- Dmitry Korobchenko, Deep Learning R&D Engineer at NVIDIA.
Tickets: https://provectus.com/datasummer/#tickets
Promocode for 10% off - MLWorldDataFriends
2 streams: BigData and DataScience
Speakers from the world's biggest companies: AWS, Spotify, GridGain Systems, Google Developers
Giuseppe Angelo Porcelli, Solutions Architect at Amazon Web Services.
- Jonathan Taws, Data Scientist at Amazon Web Services;
- Sri Sri Perangur, Senior Data Scientists at Spotify
- Javier Rodriguez Zaurin, Head of Data Science at Simply Business;
- Akmal Chaudhri, Lead Technical Evangelist at GridGain Systems;
- Jacek Laskowski, Spark & Kafka Developer and Technical Instructor at Kafka Streams Consultant;
- Fedor Navruzov, Senior Data Scientist at SWAG Speak With A Geek;
- Giorgi Jvaridze, Software Engineer at Data Science Stream;
- Roman Strochak, Computer vision CTO at Data AI;
- Rudradeb Mitra, Product Mentor at Google Developers;
- Dmitry Korobchenko, Deep Learning R&D Engineer at NVIDIA.
Tickets: https://provectus.com/datasummer/#tickets
Promocode for 10% off - MLWorldDataFriends
Interesting article about Google Transformer for automatic language translation, not very https://ai.googleblog.com/2017/08/transformer-novel-neural-network.html
research.google
Transformer: A Novel Neural Network Architecture for Language Understanding
Posted by Jakob Uszkoreit, Software Engineer, Natural Language Understanding Neural networks, in particular recurrent neural networks (RNNs), are n...
Interactive segmentation http://www.vision.ee.ethz.ch/~cvlsegmentation/dextr/
www.vision.ee.ethz.ch
Deep Extreme Cut
DEXTR
Domain Adaptation with Dirt-T https://github.com/RuiShu/dirt-t
GitHub
GitHub - RuiShu/dirt-t: A DIRT-T Approach to Unsupervised Domain Adaptation (ICLR 2018)
A DIRT-T Approach to Unsupervised Domain Adaptation (ICLR 2018) - GitHub - RuiShu/dirt-t: A DIRT-T Approach to Unsupervised Domain Adaptation (ICLR 2018)
Humans pose detection using radio wave signals http://rfpose.csail.mit.edu/