generative elements of interior decoration by richard lord

​​Salesforce opensourced AI-framework for economic RL

AI Economist is capable of learning dynamic tax policies that optimize equality along with productivity in simulated economies, outperforming alternative tax systems.

Github: https://github.com/salesforce/ai-economist
Blog post with results: https://blog.einstein.ai/the-ai-economist/
Blog post with release: https://blog.einstein.ai/the-ai-economist-moonshot/

#Salesforce #gym #RL #economics #AIEconomics #animalcrossing #AIEconomist

​​announcing scann: efficient vector similarity search
ruiqi guo, philip sun, erik lindgren, quan geng, david simcha, felix chern, & sanjiv kumar @ google research

scann is a method for efficient vector similarity search at scale. them implements includes search space pruning & quantization for maximum inner product search & also supports other distance functions such as euclidean distance
the implementation is designed for x86 processors with avx2 support
scann achieves sota performance on ann-benchmarks.com as shown on the glove-100-angular dataset on the attached


blog post: https://ai.googleblog.com/2020/07/announcing-scann-efficient-vector.html
paper: https://arxiv.org/abs/1908.10396
github: https://github.com/google-research/google-research/tree/master/scann

#icml2020 #similarity #scann #annoy

Gentle reminder that comments are available for some posts.

Click button 'Comments' and ask questions or share your opinion.

in the walls
by matt bierner

to make a scene from a horror movie when a face comes out of a wall like the 1st season of "Very Strange Things"
based on the arkit with ar & facetracking from the front camera


on the app store only: https://apps.apple.com/ru/app/in-the-walls/id1522257130?l=en

#arkit #ar #app

Caption: ArXiV paper, in Experiment, in Production

Source: https://vk.com/wall-166490678_545

Simple scalable graph neural networks

Michael Bronstein continues a marathon of great blog posts on GML. In a new post he describes their recent work on scaling GNNs to large network. There is a good introduction to sampling-based methods (e.g. SAGE, GraphSAINT, ClusterGCN), which sample a subgraph of a large graph and then train GNN only on a subgraph.

Then, he describes that it can be beneficial just precompute r-hop matrices, A^r X, and use MLP on these features. This way, you use topology of your graph and you apply mini-batch training with MLP.

What's cool is that the algorithm is already available in pytorch-geometric as a transform, which is implemented via sparseTensor matrix multiplication.

📝 Post "Simple scalable graph neural networks" published, discuss!

​​NeuralCam Live release on the #ProductHunt

App turns iPhone into the better camera for Zoom calls with auto bluring in case of unwanted gestures.
It was clear that global pandemic and pressure on the remote culture will be foundation for new ideas and solutions, such as this.
There is nothing groundbraking about technology, but execution and market is what matters. Apple or Google might even buy this startup instead of simply copying the features and making default cameras more smart.


ProductHunt: https://www.producthunt.com/posts/neuralcam-live

#aiproduct #dataproduct #camera #aicamera #cv #DL

Google AI Blog: On-device Supermarket Product Recognition http://feedproxy.google.com/~r/blogspot/gJZg/~3/uEq7NDB-AgY/on-device-supermarket-product.html

📝 Post "Google AI Blog: On-device Supermarket Product…" published, discuss!

nlp newsletter 14: nlp beyond english, big bird, monitoring ml models, breaking into nlp, arxiv dataset,…
by elvis saravia @dair.ai

in our point of view in this newsletter showcase the next interesting links
* demos and applications gpt3
* monitoring ml models
* Big Bird: Transformers for Longer Sequences by reducing the complexity of the attention mechanism to linear complexity in the number of tokens
* competition contradictory, my dear watson: detecting contradiction and entailment in multilingual text using tpus
* competition hate speech and offensive content identification in indo-european languages
* why u should do nlp beyond :en: by sebastian ruder
* covost v2: expanding the largest, most diverse multilingual speech-to-text translation data set
* panel discussion about the future of conversational ai systems
* …


blog post: https://dair.ai/NLP_Newsletter_14-en/

#nlp #news

​​REALM: Integrating Retrieval into Language Representation Models
by google research

A new paper from google with a novel approach for language model pre-training, which augments a language representation model with a knowledge retriever.

The idea is the following: we take a sentence or a piece of text and augment it with additional knowledge (pass original text and additional texts to the model).

An example:
The masked text is:

We paid twenty __ at the Buckingham Palace gift shop.


Knowledge retriever could add the following information to it:
Buckingham Palace is the London residence of the British monarchy.
The official currency of the United Kingdom is the Pound.


blog post: https://ai.googleblog.com/2020/08/realm-integrating-retrieval-into.html
paper: https://arxiv.org/abs/2002.08909
github: https://github.com/google-research/language/tree/master/language/realm

#nlp #languagemodel #knowledgeretriever #icml2020

​​A utility tool powered by fzf for using git interactively.

This tool is designed to help you use git more efficiently. It's lightweight and easy to use.

Also integrates with: diff-so-fancy, delta, bat, emoji-cli.

https://github.com/wfxr/forgit

#shell #git

The Quantum Graph Recurrent Neural Network

This demonstration by pennylane investigates quantum graph recurrent neural networks (QGRNN), which are the quantum analogue of a classical graph recurrent neural network, and a subclass of the more general quantum graph neural network ansatz. Both the QGNN and QGRNN were introduced in this paper (2019) by Google X.

​​Clothing Dataset: Call for Action

Help to collect a public-domain dataset with images of clothes

Medium post: https://medium.com/data-science-insider/clothing-dataset-call-for-action-3cad023246c1

#dataset #clothing #cv #calltoarms

​​GANs used to create photorealistic images of Roman Emperors

Project post: https://voshart.com/ROMAN-EMPEROR-PROJECT
Medium: https://medium.com/@voshart/photoreal-roman-emperor-project-236be7f06c8f

#GAN #Art #history #DL

mingpt – a minimal pytorch re-implementation of the openai generative pretrained transformer training
by karpathy

small, clean, interpretable and educational, as most of the currently available ones are a bit sprawling. this implementation is appropriately about 300 lines of code, including boilerplate and a totally unnecessary custom causal self-attention module. all that's going on is that a sequence of indices goes into a sequence of transformer blocks, and a probability distribution of the next index comes out.

with a bpe encoder, distributed training and maybe fp16 this implementation may be able to reproduce gpt-1/gpt-2 results, though they haven't tried $$$. gpt-3 is likely out of reach as his understanding is that it does not fit into gpu memory and requires a more careful model-parallel treatment.


https://twitter.com/karpathy/status/1295410274095095810?s=20

#nlp #karpathy #gpt #torch