Object Detection: EfficientDet (SOTA), MobileNetv3 and YOLO using OpenCV and TensorFlow.

- Blog post: https://imadelhanafi.com/posts/object_detection_yolo_efficientdet_mobilenet/

- Github repo: https://github.com/imadelh/Object-Detection_MobileNetv3-EfficientDet-YOLO

- Live version: https://vision.imadelhanafi.com/predict/v1?model=MODEL_NAME&image_url=URL (example: https://vision.imadelhanafi.com/predict/v1?model=mobilenet&image_url=https://imadelhanafi.com/data/draft/random/img4.jpg)

AI postdocs available! Stanford AI Lab is delighted to offer postdocs to some exciting young AI researchers in these difficult times. Positions for 2 years working with SAIL faculty. If you’ve procrastinated, this is the week to get your application in!
https://ai.stanford.edu/postdoctoral-applications/


@ArtificialIntelligenceArticles

From Ian Goodfellow and other Google researchers: A novel approach to generating high-resolution images, guided by small inputs, that results in perceptually convincing details (called Latent Adversarial Generator (LAG))

For project and code or API request: https://www.catalyzex.com/paper/arxiv:2003.02365

Yann LeCun thinks tensor networks are similar to convolutional neural networks

http://www.preposterousuniverse.com/blog/2015/05/05/does-spacetime-emerge-from-quantum-information/

A Metric Learning Reality Check
Musgrave et al.: https://arxiv.org/abs/2003.08505
"Our results show that when hyperparameters are properly tuned via cross-validation, most methods perform similarly to one another"
#ArtificialIntelligence #DeepLearning #MachineLearning

Text classification with Transformer
Apoorv Nandan, Colab : https://colab.research.google.com/github/keras-team/keras-io/blob/master/examples/nlp/ipynb/text_classification_with_transformer.ipynb
#ArtificialIntelligence #DeepLearning #Transformer

NVIDIA’s New Ampere Data Center GPU
NVIDIA A100 GPU is a 20x AI performance leap and an end-to-end machine learning accelerator : https://nvidianews.nvidia.com/news/nvidias-new-ampere-data-center-gpu-in-full-production
#NVIDIA #GPU #DeepLearning

Using Reinforcement Learning in the Algorithmic Trading Problem
Ponomarev et al.: https://arxiv.org/abs/2002.11523
#DeepLearning #ReinforcementLearning #Trading

LandCover.ai: Dataset for Automatic Mapping of Buildings, Woodlands and Water from Aerial Imagery

For project and dataset: https://www.catalyzex.com/paper/arxiv:2005.02264

They collected images of 216.27 sq. km lands across Poland, a country in Central Europe, 39.51 sq. km with resolution 50 cm per pixel and 176.76 sq. km with resolution 25 cm per pixel and manually fine annotated three following classes of objects: buildings, woodlands, and water.

Jim Keller - Moore's Law in the age of AI Chips
https://www.youtube.com/watch?v=8eT1jaHmlx8

​​CREME – python library for online ML

All the tools in the library can be updated with a single observation at a time, and can therefore be used to learn from streaming data.

The model learns from one observation at a time, and can therefore be updated on the fly. This allows to learn from massive datasets that don't fit in main memory. Online machine learning also integrates nicely in cases where new data is constantly arriving. It shines in many use cases, such as time series forecasting, spam filtering, recommender systems, CTR prediction, and IoT applications. If you're bored with retraining models and want to instead build dynamic models, then online machine learning might be what you're looking for.

Here are some benefits of using creme (and online machine learning in general):
• incremental – models can update themselves in real-time
• adaptive – models can adapt to concept drift
• production-ready – working with data streams makes it simple to replicate production scenarios during model development
• efficient – models don't have to be retrained and require little compute power, which lowers their carbon footprint


api reference: https://creme-ml.github.io/content/api.html
github: https://github.com/creme-ml/creme

#ml #online #learning

Free Certification Course on Deep Learning with PyTorch in partnership with freeCodeCamp


https://docs.google.com/forms/d/e/1FAIpQLSeFj1h6Z8mtedlg0i3alB0NE5-ECBmIhUVNw53RLGEd8QF8Vg/viewform

Teaching from Home - Quick Start Guide

By Andrew Ng

Many of us are working to quickly transition from teaching in a live classroom to teaching online
from home. The goal of this document is to help you make that transition quickly and
successfully with a minimum amount of complexity. We will go over the basics, and only the
basics here.
@ArtificialIntelligenceArticles

https://drive.google.com/file/d/1ZPUQTKxkMLPxinT4SHU3_k_p4_Scnqgv/view

@ArtificialIntelligenceArticles

Blog: https://arstechnica.com/gaming/2020/05/after-watching-50000-hours-of-pac-man-nvidias-ai-generated-a-playable-clone/

Video: https://youtu.be/BYt6r8z6pUY

Missing Semester

As computer scientists, we know that computers are great at aiding in repetitive tasks. However, far too often, we forget that this applies just as much to our use of the computer as it does to the computations we want our programs to perform. We have a vast range of tools available at our fingertips that enable us to be more productive and solve more complex problems when working on any computer-related problem. Yet many of us utilize only a small fraction of those tools; we only know enough magical incantations by rote to get by, and blindly copy-paste commands from the internet when we get stuck.

This class is an attempt to address this.

We want to teach you how to make the most of the tools you know, show you new tools to add to your toolbox, and hopefully instill in you some excitement for exploring (and perhaps building) more tools on your own. This is what we believe to be the missing semester from most Computer Science curriculum

https://www.youtube.com/watch?v=Z56Jmr9Z34Q&list=PLyzOVJj3bHQuloKGG59rS43e29ro7I57J

MediaPipe Hand
MediaPipe Hand is a high-fidelity hand and finger tracking solution. GitHub : https://github.com/google/mediapipe
#DeepLearning #MachineLearning #MediaPipe