HATS: A NEW EVENT-BASED  OBJECT CLASSIFICATION METHOD THAT IS 39x FASTER THAN ITS CLOSEST COMPETITOR

HATS paper from CVPR

A paper recently published by Prophesee’s team presents HATS: Histograms of Averaged Time Surfaces for Robust Event-based Object Classification.

The paper was selected to be presented at CVPR 2018, a highly prestigious and selective conference.

There were 2 key motivations behind the paper :

-1

The lack of  low-level representations and architectures for event-based sensors

HATS use a new event-based feature representation and a new machine learning architecture for object classification, described in the poster below.

These are designed to take advantage of the higher temporal resolution, and local memory of event-based sensors.

%

Accuracy using a linear classifier

faster than CSNN

-2

The absence of large real-world event-based datasets

HATS was released along with the first large real-world event-based dataset for object classification: N-CARS.

HATS achieves 90% accuracy for car classification results using a linear classifier, 39x faster than convolutional spiking neural networks – the closest competitor. These results are shown in figures in the poster below

The paper is summarized in the poster below, which was selected to be presented at CVPR 2018

Slide The first industrial grade Event-Based Vision system Releasing new Robot Operating System driver What is Event-Based Vision ? PROPHESEE wins inVision Top Innovation award