🔍 Advancing computer vision research with new Detectron2 Mask R-CNN baselines

Facebook Ai: https://ai.facebook.com/blog/advancing-computer-vision-research-with-new-detectron2-mask-r-cnn-baselines/

Code: https://github.com/facebookresearch/detectron2/blob/master/MODEL_ZOO.md#new-baselines-using-large-scale-jitter-and-longer-training-schedule

Tensorflow implementation: https://github.com/tensorflow/tpu/tree/master/models/official/detectiona

@ai_machinelearning_big_data

📘 D2L.ai: Interactive Deep Learning Book with Multi-Framework Code, Math, and Discussions

Github: https://github.com/d2l-ai/d2l-en

Book: https://d2l.ai/

Paper: https://arxiv.org/abs/2106.11342v1

@ai_machinelearning_big_data

Cartoon-StyleGan2 🙃 : Fine-tuning StyleGAN2 for Cartoon Face Generation

Github: https://github.com/happy-jihye/Cartoon-StyleGan2

Paper: https://arxiv.org/abs/2106.12445

Colab: https://colab.research.google.com/github/happy-jihye/Cartoon-StyleGan2/blob/main/Cartoon_StyleGAN2.ipynb

@ai_machinelearning_big_data

🔺 Pyramid Vision Transformer

Image classification, object detection, and semantic segmentation tasks

Github: https://github.com/whai362/PVT

Paper: https://arxiv.org/abs/2106.13797v2

@ai_machinelearning_big_data

🔎 Microsoft AutoML - Neural Architecture Search

New one-shot architecture search framework dedicated to vision transformer search

Github: https://github.com/microsoft/AutoML

Paper: https://arxiv.org/abs/2107.00651v1

Models: https://drive.google.com/drive/folders/1NLGAbBF9bA1IUAxKlk2VjgRXhr6RHvRW

Dataset: https://paperswithcode.com/dataset/cifar-10

@ai_machinelearning_big_data

🚀 TensorFlow and PyTorch performance benchmarking in 2021

Habr: https://habr.com/ru/company/ru_mts/blog/565456/

Github: https://github.com/Chifffa/tf_vs_torch_benchmarking

@ai_machinelearning_big_data

🔝 Learning Hierarchical Graph Neural Networks for Image Clustering

Github: https://github.com/dmlc/dgl/tree/master/examples/pytorch/hilander

Paper: https://arxiv.org/abs/2107.01319

Datasets: https://drive.google.com/file/d/1KLa3uu9ndaCc7YjnSVRLHpcJVMSz868v/view

@ai_machinelearning_big_data

🌐 Depth-supervised NeRF: Fewer Views and Faster Training for Free

Github: https://github.com/dunbar12138/DSNeRF

Paper: https://arxiv.org/abs/2107.02791v1

Project: https://www.cs.cmu.edu/~dsnerf/

Datasets: https://paperswithcode.com/dataset/3dmatch

@ai_machinelearning_big_data

↔️ Independent Encoder for Deep Hierarchical Unsupervised Image-to-Image Translation

Github: https://github.com/Elvinky/IEGAN

Paper: https://arxiv.org/abs/2107.02494

Datasets: https://github.com/Elvinky/IEGAN/tree/main/dataset

Image-to-Image Translation: https://paperswithcode.com/task/image-to-image-translation

@ai_machinelearning_big_data

📐 Googl's Measuring and Improving Model-Moderator Collaboration using Uncertainty Estimation

Providing high-quality implementations of standard and state-of-the-art methods on standard tasks.

Github: https://github.com/google/uncertainty-baselines

Paper: https://arxiv.org/abs/2107.04212v1

Dataset: https://www.tensorflow.org/datasets

@ai_machinelearning_big_data

⚛ Highly accurate protein structure prediction with AlphaFold

https://deepmind.com/blog/article/putting-the-power-of-alphafold-into-the-worlds-hands

Github: https://github.com/deepmind/alphafold

Paper: https://doi.org/10.1038/s41586-021-03819-2

@ai_machinelearning_big_data

👨‍🎓 From economists to data scientists or how to become the leader of the Kaggle Notebooks rating

Habr: https://habr.com/ru/company/ru_mts/blog/567678/

Exploration of data step by step: https://www.kaggle.com/artgor/exploration-of-data-step-by-step

@ai_machinelearning_big_data

📖 GFP-GAN: Towards Real-World Blind Face Restoration with Generative Facial Prior

Github: https://github.com/TencentARC/GFPGAN

Paper: https://arxiv.org/abs/2101.04061v2

Dataset: https://paperswithcode.com/dataset/lfw

Colab Demo: https://colab.research.google.com/drive/1sVsoBd9AjckIXThgtZhGrHRfFI6UUYOo

@ai_machinelearning_big_data

🐍 PyTorch Fundamentals Free Microsoft Course

https://docs.microsoft.com/en-us/learn/paths/pytorch-fundamentals/

Ru: https://docs.microsoft.com/ru-ru/learn/paths/pytorch-fundamentals/

Github: https://github.com/pytorch/tutorials

@ai_machinelearning_big_data

🔝 Deepmind's WikiGraphs: A Wikipedia Text - Knowledge Graph Paired Dataset

This package provides tools to download the WikiGraphs dataset

Github: https://github.com/deepmind/deepmind-research/tree/master/wikigraphs

Paper: https://arxiv.org/abs/2107.09556v1

Dataset: https://paperswithcode.com/dataset/wikigraphs

@ai_machinelearning_big_data

Best machine learning tutorials: https://t.iss.one/datascienceiot

Usufull python resourses: @pythonl

Artificial intelligence articles: @ArtificialIntelligencedl

Machine learning RU: https://t.iss.one/machinelearning_ru

ML chat: https://t.iss.one/machinee_learning

Free python books: https://t.iss.one/pythonlbooks

🌀 CycleMLP: A MLP-like Architecture for Dense Prediction

Github: https://github.com/ShoufaChen/CycleMLP

Paper: https://arxiv.org/abs/2107.10224

Dataset: https://paperswithcode.com/dataset/imagenet

@ai_machinelearning_big_data

Тут у Яндекса интересная новость. Компания запускает соревнование для исследователей в области машинного обучения в рамках крупнейшей конференции MLщиков в мире - NeurIPS 2021. Вместе с учеными Оксфорда и Кембриджа предлагают участникам посоревноваться в разработке алгоритмов и их обучении для погоды, машинного перевода текстов и предсказания поведения участников автомобильного движения. Основной задачей будет проверить эффективность этих алгоритмов при сдвиге данных.

Для соревнования Яндекс открыл доступ к собственному датасету, который считается самым большим в мире по беспилотным автомобилям. Еще поделятся реальными данными Я.Погоды и Я.Переводчика. Это данные из сервисов, которые много лет работают в реальном мире, используются в различных сценариях, и уже проходили испытание сдвигом данных.

Полученные решения можно будет применять в разных отраслях, которые сталкиваются со сдвигом данных. Крутая инициатива!

https://research.yandex.com/shifts

🤗 The Best SOTA NLP Course is Free!

https://huggingface.co/course/chapter1

Dataset: https://github.com/huggingface/datasets

Colab: https://colab.research.google.com/github/huggingface/notebooks/blob/master/course/chapter1/section3.ipynb

@ai_machinelearning_big_data

https://t.iss.one/Golang_google - golang channel

https://t.iss.one/javascriptv - javascript tutorials

https://t.iss.one/memes_prog - it memes

https://t.iss.one/pro_python_code - python ru

https://t.iss.one/htmlcssjavas - web development

https://t.iss.one/csharp_ci - c sharp

https://t.iss.one/linuxkalii - kali linux