Understanding the Role of Individual Units in a Deep Neural Network

Examine two types of networks that contain interpretable units: networks trained to classify images of scenes, and networks trained to synthesize images of scenes.

https://dissect.csail.mit.edu/

Github: https://github.com/davidbau/dissect

Website: https://www.pnas.org/content/early/2020/08/31/1907375117

Paper: https://arxiv.org/pdf/2009.05041.pdf

Up Great technology contest READ//ABLE stimulates the development of new approaches to machine learning. It gives great opportunities for NLP developers. Join us for the next AI breakthrough!

Task: to develop an AI product capable of successfully identifying semantic and factual errors in academic essays at the specialist level within the limited time.

Prize: 100 million rubles for each language: Russian or English

Dataset: https://bit.ly/34279IC

A set of text essay files in Russian and English is a usable tool for specialists in the field. The dataset will be replenished.

Info and terms of participation: https://bit.ly/3kGdTBZ

LaSOT

Large-scale Single Object Tracking (LaSOT) aims to provide a dedicated platform for training data-hungry deep trackers as well as assessing long-term tracking performance.

http://vision.cs.stonybrook.edu/~lasot/

Github: https://github.com/HengLan/LaSOT_Evaluation_Toolkit

Dataset: http://vision.cs.stonybrook.edu/~lasot/download.html

Paper: https://arxiv.org/abs/2009.03465

@ai_machinelearning_big_data

Rule-Guided Graph Neural Networks for Recommender Systems

Сombination of rule learning and GNNs achieves substantial improvement compared to methods only using either of them

Github: https://github.com/nju-websoft/RGRec

Paper: https://arxiv.org/abs/2009.04104v1

Improving Sparse Training with RigL

https://ai.googleblog.com/2020/09/improving-sparse-training-with-rigl.html

Github: https://github.com/google-research/rigl

Paper: https://arxiv.org/abs/1911.11134

@ai_machinelearning_big_data

Dialog Ranking Pretrained Transformers

It is a set of dialog response ranking models proposed by Microsoft Research NLP Group trained on millions of human feedback data.

Github: https://github.com/golsun/DialogRPT

Paper: https://arxiv.org/abs/2009.06978

Colab: https://colab.research.google.com/drive/1jQXzTYsgdZIQjJKrX4g3CP0_PGCeVU3C?usp=sharing

@ai_machinelearning_big_data

MEAL V2

Boosting Vanilla ResNet-50 to 80%+ Top-1 Accuracy on ImageNet without Tricks.

Github: https://github.com/szq0214/MEAL-V2

Paper: https://arxiv.org/abs/2009.08453

ImageNet dataset: https://github.com/pytorch/examples/tree/master/imagenet#requirements.

@ai_machinelearning_big_data

Advancing NLP with Efficient Projection-Based Model Architectures

https://ai.googleblog.com/2020/09/advancing-nlp-with-efficient-projection.html

Sequence Projection Models: https://github.com/tensorflow/models/tree/master/research/sequence_projection

📸 Old Photo Restoration (Official PyTorch Implementation)

Restore old photos that suffer from severe degradation through a deep learning approace.

http://raywzy.com/Old_Photo/

Github: https://github.com/microsoft/Bringing-Old-Photos-Back-to-Life

Paper: https://arxiv.org/pdf/2009.07047v1.pdf

Colab: https://colab.research.google.com/drive/1NEm6AsybIiC5TwTU_4DqDkQO0nFRB-uA

@ai_machinelearning_big_data

Facebook AI Releases ‘Dynabench’, A Dynamic Benchmark Testing Platform For Machine Learning Systems

Articel: https://ai.facebook.com/blog/dynabench-rethinking-ai-benchmarking/

Project: https://dynabench.org/

@ai_machinelearning_big_data

Bringing the Mona Lisa Effect to Life with TensorFlow.js

https://blog.tensorflow.org/2020/09/bringing-mona-lisa-effect-to-life-tensorflow-js.html

Github: https://github.com/emilyxxie/mona_lisa_eyes

Demo: https://monalisaeffect.com/

@ai_machinelearning_big_data

🔋 The Most Complete Guide to PyTorch for Data Scientists

https://www.kdnuggets.com/2020/09/most-complete-guide-pytorch-data-scientists.html

Code: https://github.com/MLWhiz/data_science_blogs/tree/master/pytorch_guide

@ai_machinelearning_big_data

Graph Normalization

Learning Graph Normalization for Graph Neural Networks

Github: https://github.com/cyh1112/GraphNormalization

Paper: https://arxiv.org/abs/2009.11746v1

@ai_machinelearning_big_data

CaGNet: Context-aware Feature Generation for Zero-shot Semantic Segmentation.

Github: https://github.com/bcmi/CaGNet-Zero-Shot-Semantic-Segmentation

Paper: https://arxiv.org/abs/2009.12232v1

@ai_machinelearning_big_data

Seeing Theory

🎲 A visual introduction to probability and statistics

https://seeing-theory.brown.edu/index.html#4thPage

📗 Free book: https://seeing-theory.brown.edu/doc/seeing-theory.pdf

Utterance-level Dialogue Understanding: An Empirical Study

The recent abundance of conversational data on the Web and elsewhere calls for effective NLP systems for dialog understanding.

Github: https://github.com/declare-lab/conv-emotion

Paper: https://arxiv.org/abs/2009.13902v1

Transfer learning with TensorFlow Hub  |  TensorFlow Core

https://www.tensorflow.org/tutorials/images/transfer_learning_with_hub

TensorFlow Hub : https://tfhub.dev/

Github: https://github.com/tensorflow/docs/blob/master/site/en/tutorials/images/transfer_learning_with_hub.ipynb

@tensorflowblog

Rotated Binary Neural Network

Pytorch implementation of RBNN.

Github: https://github.com/lmbxmu/RBNN

Paper: https://arxiv.org/abs/2009.13055

@ai_machinelearning_big_data

aLRP Loss: A Ranking-based, Balanced Loss Function

Unifying Classification and Localisation in Object Detection.

💻 Github: https://github.com/kemaloksuz/aLRPLoss

📎 Dataset: https://cocodataset.org/#download

🗒 Paper: https://arxiv.org/abs/2009.13592v1

@ai_machinelearning_big_data