Preserving Detail in Image Inputs: Better image compression for computer vision datasets
Vision

Preserving Detail in Image Inputs: Better image compression for computer vision datasets

Given real-world constraints on memory and processing time, images are often downsampled before they’re fed into a neural network. But the process removes fine details, and that degrades accuracy. A new technique squeezes images with less compromise.

April 22, 20202 min read
A Neural Net in Every Bathroom: A smart toilet uses AI to monitor waste for signs of disease.
Vision

A Neural Net in Every Bathroom: A smart toilet uses AI to monitor waste for signs of disease.

It’s time to stop flushing valuable data down the toilet. The Precision Health Toilet, a suite of sensors that attach to an ordinary commode, monitors human waste for input for signs of disease. It identifies individual users by scanning where the sun doesn’t shine.

April 22, 20202 min read
Preserving Detail in Image Inputs: Better image compression for computer vision datasets
Vision

Preserving Detail in Image Inputs: Better image compression for computer vision datasets

Given real-world constraints on memory and processing time, images are often downsampled before they’re fed into a neural network. But the process removes fine details, and that degrades accuracy. A new technique squeezes images with less compromise.

April 22, 20202 min read
Workers of the World, Don’t Unite: Computer vision helped workers maintain social distance.
Vision

Workers of the World, Don’t Unite: Computer vision helped workers maintain social distance.

Computer vision is helping construction workers keep their social distance. Smartvid.io, a service that focuses on construction sites, offers a tool that recognizes when workers get too close to each other. The tool sends distancing warnings and reports to construction superintendents.

April 22, 20201 min read
Outside the Norm: Batch normalization contributes to neural network accuracy.
Vision

Outside the Norm: Batch normalization contributes to neural network accuracy.

Batch normalization is a technique that normalizes layer outputs to accelerate neural network training. But new research shows that it has other effects that may be more important.

April 8, 20202 min read
Outside the Norm: Batch normalization contributes to neural network accuracy.
Vision

Outside the Norm: Batch normalization contributes to neural network accuracy.

Batch normalization is a technique that normalizes layer outputs to accelerate neural network training. But new research shows that it has other effects that may be more important.

April 8, 20202 min read
Beyond Neural Architecture Search: AutoML-Zero is a meta-algorithm for classification.
Vision

Beyond Neural Architecture Search: AutoML-Zero is a meta-algorithm for classification.

Faced with a classification task, an important step is to browse the catalog of machine learning architectures to find a good performer. Researchers are exploring ways to do it automatically.

April 8, 20202 min read
Beyond Neural Architecture Search: AutoML-Zero is a meta-algorithm for classification.
Vision

Beyond Neural Architecture Search: AutoML-Zero is a meta-algorithm for classification.

Faced with a classification task, an important step is to browse the catalog of machine learning architectures to find a good performer. Researchers are exploring ways to do it automatically.

April 8, 20202 min read
(Science) Community Outreach: A survey of machine learning from Eric Schmidt
Vision

(Science) Community Outreach: A survey of machine learning from Eric Schmidt

Are your scientist friends intimidated by machine learning? They might be inspired by a primer from one of the world’s premier tech titans. Former Google CEO Eric Schmidt and Cornell PhD candidate Maithra Raghu school scientists in machine learning in a sprawling overview.

April 8, 20201 min read
(Science) Community Outreach: A survey of machine learning from Eric Schmidt
Vision

(Science) Community Outreach: A survey of machine learning from Eric Schmidt

Are your scientist friends intimidated by machine learning? They might be inspired by a primer from one of the world’s premier tech titans. Former Google CEO Eric Schmidt and Cornell PhD candidate Maithra Raghu school scientists in machine learning in a sprawling overview.

April 8, 20201 min read
Where Are the Live Bombs?: Computer vision identifies unexploded bombs in Cambodia.
Vision

Where Are the Live Bombs?: Computer vision identifies unexploded bombs in Cambodia.

Unexploded munitions from past wars continue to kill and maim thousands of people every year. Computer vision is helping researchers figure out where these dormant weapons are likely to be.

April 1, 20202 min read
Where Are the Live Bombs?: Computer vision identifies unexploded bombs in Cambodia.
Vision

Where Are the Live Bombs?: Computer vision identifies unexploded bombs in Cambodia.

Unexploded munitions from past wars continue to kill and maim thousands of people every year. Computer vision is helping researchers figure out where these dormant weapons are likely to be.

April 1, 20202 min read
Seeing the See-Through: ClearGrasp allows robots to grab see-through objects.
Vision

Seeing the See-Through: ClearGrasp allows robots to grab see-through objects.

Glass bottles and crystal bowls bend light in strange ways. Image processing networks often struggle to separate the boundaries of transparent objects from the background that shows through them. A new method sees such items more accurately.

April 1, 20202 min read
Seeing the See-Through: ClearGrasp allows robots to grab see-through objects.
Vision

Seeing the See-Through: ClearGrasp allows robots to grab see-through objects.

Glass bottles and crystal bowls bend light in strange ways. Image processing networks often struggle to separate the boundaries of transparent objects from the background that shows through them. A new method sees such items more accurately.

April 1, 20202 min read
Rightsizing Neural Nets: An equation for predicting optimal data and model size
Vision

Rightsizing Neural Nets: An equation for predicting optimal data and model size

How much data do we want? More! How large should the model be? Bigger! How much more and how much bigger? New research estimates the impact of dataset and model sizes on neural network performance.

March 25, 20202 min read

Subscribe to The Batch

Stay updated with weekly AI News and Insights delivered to your inbox