Artificial Intelligence: Week #3 | 2021

This week in AI & Machine Learning: Tracking Elephants, Tensorflow on Raspberry Pi Pico, detection in the browser, On-Device ML study jam, and more

Artificial Intelligence News:

Conservationists Track Elephant Populations from Space Using Computer Vision 

If you’ve been reading these weekly AI news segments, you’ll know this is far from the first time we’ve seen artificial intelligence used to help elephants! Researchers are now using satellites and computer vision to track vulnerable elephant population more effectively and efficiently. 

Machine Learning Could Cut Delays from Traffic Lights

Learn how cities in the future could end up using machine learning for traffic light control to minimize driving delays. It’s fascinating to think about how cities may change in the future with more connected autonomous vehicles and devices.  

Developer Tools & Education:

Meet Raspberry Silicon: Raspberry Pi Pico now on sale at $4

The Raspberry Pi Pico is now available to buy for only $4! This microprocessor board seems really neat, and it will likely receive wide support and adoption like other Raspberry Pi products. The TensorFlow team is already working on a port of TFlite Micro for it!

Contrastive Loss for Siamese Networks with Keras and TensorFlow

In this week’s post from pyimagesearch you will learn about contrastive loss and how it can be used to train more accurate siamese neural networks with Keras and TensorFlow.

Custom object detection in the browser using TensorFlow.js

Learn how to build your own custom object detection model to run in the browser using TensorFlow.js. 

How to Choose an Activation Function for Deep Learning

In this post from Machine Learning Mastery, learn about choosing an activation function for deep learning models.  

Interesting Podcasts & Interviews:

Max Tegmark: AI and Physics | Lex Fridman

Max Tegmark, physicist, AI researcher at MIT, and author of life 3.0 talks about AI safety, physics, autonomous weapons, AI alignment, and much more.

Otter.ai CEO Sam Liang on Bringing Live Captions to a Meeting Near You

Sam Liang, CEO and co-founder of otter.ai talks about using AI to produce speech-to-text transcriptions in real time for video streaming platforms.

AI for Ecology and Ecosystem Preservation with Bryan Carstens

Bryan Carstens, a professor in the Department of Evolution, Ecology, and Organismal Biology tackles understanding for the vast known unknowns in species and biodiversity, and how he came to apply machine learning to his lab’s research.

Off-Line, Off-Policy RL for Real-World Decision Making at Facebook

Jason Gauci, a Software Engineering Manager at Facebook AI discusses their Reinforcement Learning platform, Re-Agent (Horizon) and more.

Notable Research Papers:

• Does Continual Learning = Catastrophic Forgetting?

• Learn to Dance with AIST++: Music Conditioned 3D Dance Generation

• Interpretable Models for Granger Causality Using Self-explaining Neural Networks

• IntentNet: Learning to Predict Intention from Raw Sensor Data

• Disentangled Recurrent Wasserstein Autoencoder

• Cross-modal Learning for Domain Adaptation in 3D Semantic Segmentation

• Multimodal Variational Autoencoders for Semi-Supervised Learning: In Defense of Product-of-Experts