ImageNet Large Scale Visual Recognition Challenge

TL;DR

The ImageNet Large Scale Visual Recognition Challenge is a benchmark for object classification and detection that has driven significant advances in computer vision over the past decade.

Contribution

This paper introduces the ImageNet benchmark dataset and analyzes key progress and challenges in large-scale image recognition and detection.

Findings

Significant improvements in object recognition accuracy

Comparison of computer vision performance with human accuracy

Lessons learned and future directions for large-scale visual recognition

Abstract

The ImageNet Large Scale Visual Recognition Challenge is a benchmark in object category classification and detection on hundreds of object categories and millions of images. The challenge has been run annually from 2010 to present, attracting participation from more than fifty institutions. This paper describes the creation of this benchmark dataset and the advances in object recognition that have been possible as a result. We discuss the challenges of collecting large-scale ground truth annotation, highlight key breakthroughs in categorical object recognition, provide a detailed analysis of the current state of the field of large-scale image classification and object detection, and compare the state-of-the-art computer vision accuracy with human accuracy. We conclude with lessons learned in the five years of the challenge, and propose future directions and improvements.