Fast, Accurate Detection of 100,000 Object Classes on a Single Machine a prizewinning paper by Google Research scientists, describes a breakthrough in machine vision that can distinguish between a huge class of objects 20,000 times faster than before.
This so-called convolution operator is one of the key operations used in computer vision and, more broadly, all of signal processing. Unfortunately, it is computationally expensive and hence researchers use it sparingly or employ exotic SIMD hardware like GPUs and FPGAs to mitigate the computational cost. We turn things on their head by showing how one can use fast table lookup — a method called hashing — to trade time for space, replacing the computationally-expensive inner loop of the convolution operator — a sequence of multiplications and additions — required for performing millions of convolutions with a single table lookup.
We demonstrate the advantages of our approach by scaling object detection from the current state of the art involving several hundred or at most a few thousand of object categories to 100,000 categories requiring what would amount to more than a million convolutions. Moreover, our demonstration was carried out on a single commodity computer requiring only a few seconds for each image. The basic technology is used in several pieces of Google infrastructure and can be applied to problems outside of computer vision such as auditory signal processing.
Fast, Accurate Detection of 100,000 Object Classes on a Single Machine
(Image: Clutter, a Creative Commons Attribution Share-Alike (2.0) image from neofob's photostream)
Correlation is not causation, and the data-set is awfully small (39 incidents), but computational epidemiologist Maimuna Majumder is working with what’s available, because the federal government won’t fund research into gun fatalities, and does not require states to gather data on police use of force.
In Neural Photo Editing With Introspective Adversarial Networks, a group of University of Edinburgh engineers and a private research colleague describe a method for using “introspective adversarial networks” to edit images in realtime, which they demonstrate in an open project called “Neural Photo Editor” that “enhances” photos by predicting what should be under your brush.
A trio of scholars who study the psychology and philosophy of science have written a fantastic paper for Springer’s Sythese looking at the way that climate change conspiracy theorists construct their view of the world, and how these conspiracy theories contain self-contradictory theses (like the idea that climate change can’t be predicted and the idea […]
#1. A-Audio Legacy Noise Cancelling Headphones with 3-Stage Technology The A-Audio Legacy Headphones are the Boing Boing Store’s best seller this month, and it’s easy to see why. With 40mm drivers, powerful circuitry, and memory foam padded circumaural ear cups, these are clearly super high-quality headphones. Plus, the patented 3-Stage Technology lets you toggle between passive […]
Vaping is getting more mainstream by the day, which means there’s been an influx of quality yet affordable vaporizers on the market. We’re particularly excited about the APX Wax Vaporizer Kit, which is an easy-to-use, high-quality vape that works with both dry herbs and waxy concentrates.If you’re a beginner trying to get into vaping, the APX […]
When you’ve had a long day and it’s time to unwind, there’s a lot you can do to relax: drink some tea, take a shower or even read a book. But there’s one thing that’s essential to a comfortable night’s rest—and that’s investing in some really good sheets. Enter Bamboo Bed Sheets. These quality sheets retail for $120, but […]