A summary of recent happenings in the field of Artificial Intelligence (AI) for week 35 of 2020. For more regular updates and related discussions, join us on the Artificial Intelligence Community. As always thanks for your support for sharing it with your friends and colleagues on social media.

Industry updates for week 35 of 2020

This section is for recent industry updates related to Artificial Intelligence.

  • Baidu’s Chinese NLP datasets: On its tenth anniversary, Baidu is launching a “Qian Yan” plan of 100 past Chinese NLP datasets. This might remove barriers in developing language and machine learning algorithms due to lack of datasets. 
  • AI for the discovery of new planets: Researchers from the UK have used AI to find almost 50 new planets from NASA old data stacks. This is a breakthrough in the field of Astronomy and Physics. To read more click here.
  • Demo of Elon Musk’s Neurolink brain chip: This mysterious startup is owned by Elon Musk and presented a demo on AI brain implant chip. This is certainly a breakthrough in augmenting human intelligence to machines. To read more visit this page
  • AI won against human in DARPA tournament: After a recent AlphaDogfight tournament held between AI and human, DARPA’s AI won the challenge opening many doors for AI in the field of US Military. To read more, visit here.
  • False Positive in Facial Recognition: Recently, an African-American man was arrested for a crime he did not commit because of misclassification in facial recognition software and made fool out of police after the fingerprints and DNA didn’t match the crime site. Above all, the video footage clearly showed different individual clearing out the confusion that the Machine is biased for wrong identification. To read more, visit this page
  •  Google’s Dataset search engine: After the well-established web search engine, this new dataset search has more than 30 Million datasets across 4,600 domains. Here is an article discussing complete analysis of the search engine. 
  • Verta, a machine learning platform: MIT CSAIL grad alumni’s ML platform called Verta helps researchers track model versions and operationalize them into production. It is an Intel Capital venture with participation from General Catalyst as well. 
  • Google Career Certificates: Google is making up the news a lot through its high-end research and now also due to high-end programs boosting academics along with work. To know more visit here
  • XYZ Robotics closes its piece picking series: XYZ Robotics Inc. has decided to shut down its arm-tooling for logistics and switch to AI-driven piece picking to educe hardware-driven. Read more here.
  • BabbleLabs acquired by Cisco: To remove background noise anTopeech during meetings, Cisco has decided to acquire BabbleLabs, a speech-based startup. This startup enhances human speech using AI and to know more details visit here.
  • White House finds AI interesting: AI has managed to catch the eye of world’s largest economy governed by White House. Hence, they have decided to establish 12 research institutes based on AI and quantum sciences. To read more behind the scene story, visit here.
  • AI-powered Vacuum cleanersiRobot is upgrading it’s robotic vacuum cleaners to AI-based. The vacuum bots simply would work on commands and be at your service just like the house-hold maids. 
  • Deep Learning for Banking: Visa Inc. has developed a powerful AI mechanism that allows or denies transactions in case the bank’s network is down. 
  • Automating conversations by Vernacular.ai: This AI-based SaaS business startup has been developing an AI-driven voice automation platform for transcribing long productive as well as non-productive calls over voice. The firm recently raised a 1.5milllion series A investment. To visit the company profile visit here and to read more on its investment news visit here

Articles for week 35 of 2020

This section is for recent Blog Posts/Tutorials related to Artificial Intelligence.

  • OpenBot: A project involving smartphones acting as a brain for low-budget robots. The GitHub script, article, and demo are available for reference. 
  • DETR object detection: An end-to-end object detection algorithm using transformers is described here. The Github link and article links are also available. 
  • Sentiment Classification using BERT: A nice article describes BERT for Sentiment analysis on the IMDB dataset. The technique is used to classify sentiments based on movie reviews from the dataset. To read more visit here
  • OCR Tutorial: This is a tutorial on OCR handwriting recognition using OpenCV, Keras, and Tensorflow. Visit the website here
  • Meta-Sim2: Unsupervised Learning of Scene Structure for Synthetic Data Generation, to read more on this, visit here
  • Batch Normalisation: A short course on Batch Normalisation with scripts is a must-visit in case you are new to ML and AI. 
  • Being Invisible: In case, you wish to become a little imaginative during this pandemic, here is a blog post that helps you write your script to become invisible. All you need is OpenCV and mind it because it works in real-time as well. 
  • Robotic Vision in Agriculture: The author of this post discusses various difficulties such as lighting, illumination, weather conditions and describes the classification and segmentation approach to overcomes them. To read more visit here
  • Philosopher AI: The new Philosopher AI has answers to most of your meaningless questions. The technique used is OpenAI’s GPT-3. The platform contemplates common existential questions quite well. Visit this page for the first-hand experience. 
  • Hyper-parameter Tuning: A tutorial for easy hyperparameter tuning without using any script or mathematical functions. To read more, click here
  • ML stuff that schools omit: If you are an ML enthusiast, then you would have surely visited the top free courses available and yet must have missed a few essentials that are a part of real-world scenarios. A must-read if you wish to be well prepared for any practical ML problem. 

Softwares

This section is for recent Tools/Applications related to Artificial Intelligence.

  • TF-Coder: Tensorflow expressions made easy using TF-Coder. Let’s just say we can use examples to illustrate any ML problem and get the corresponding code automatically. Read the full post here.
  • Lingvanex API: A business solution in the market to translate using neural nets, speech-to-text conversions, OCR, etc. their API can be accessed here
  • MiniTorch: A simple re-implementation of Torch API for cost efficiency and teaching purpose. It is merely a DIY and not a full-fledged library API.

Videos

This section is for recent Lectures/Conferences/Webinars related to Artificial Intelligence.

  • Workshop on Data Analytics and Machine Learning: To register visit this page
  •  Machine Folk Music School: First-ever Machine Folk Music Festival is due to be held on 13 September 2020. To register, visit this page.
  • CX-Summit 2020: Turning obstacles into opportunities. To register click here
  • AI for materials: Call for Posters, submit abstracts here
  • Github Actions with GPU: This is a tutorial video to set up GPU as a CSML docker container to execute GitHub Actions and GitLab CI workflows. 
  • Pneumonia Detection using Deep Learning: This is a project demo build using a transfer learning technique called VGG16, TensorFlow, and Keras. To watch the demo visit here
  • This is how you remove Shadows: This video presentation by Two Minute Papers portraying how images can be cleared of shadows using Machine Learning and AI. 
  • TecoGAN: This super-resolution technique is discussed by Two Minute Papers applicable for video generation. Watch the full video here
  • Fast reinforcement learning: A paper presentation involving Reinforcement learning and exploring the aspect of reduced data to solve the decision-making task is explained in this video
  • Seaborn-2: Python’s visualization library Seaborn has heatmaps and these can be seen with a correlation analysis example in this video
  • Axial-DeepLab: This video discussed the scope of this full attentional model for images. It shows promising results and is believed to be better than state-of-the-art CNNs. 

Research

This section is for recent research published related to Artificial Intelligence.

  • Funniest in FRIENDS: Using AI to find the funniest friend in the 90’s comedy series FRIENDS is an epic pandemic project. To see the project results visit this page and find more about this technique here
  • Generative Variational Autoencoder: The novel neural network called GVA is used to generate High-resolution images and the full article is accessible here. The code can be requested here
  • AI-based COVID-19 prediction model: The students at Caltech are developing a tool to predict the impact of COVID-19 using AI. They are also comparing their performance with other available models and the model can be viewed here
  • BERT-like Self-supervised models: The PyTorch repository is based on Multimodal Emotion Recognition with pre-trained Roberta and Speech-BERT. 

Miscellaneous

This section is for updates for other various related topics to Artificial Intelligence.

That’s all for the week 35 of 2020. In case you think we have missed out something and that is significant to add, please free to reach out and submit to us via our Contact form or via Social media channels: Twitter and Facebook.