Milestone · June 2009
Placed by the contemporary primary publication. The exact event date is not known; its documented interval appears below.
ImageNet provided a large dataset for visual recognition.
A shared benchmark made progress in image recognition easier to measure and compare.
A dataset’s labels and collection choices can also shape what a benchmark measures.
ImageNet · Published June 2009
Both records are about an open contest over a dataset as a way to move a field.
The AlexNet paper trains and evaluates its network on ImageNet images: direct use of the earlier dataset.
ImageNet ClassificationThe same move ImageNet later repeated: a shared labelled set as the condition for progress.
Web scale as a source of data is the precondition later datasets depended on.
The same thesis, embodied in a dataset rather than a paper.
The dataset was turned into an annual competition with a fixed metric.
Large Scale Visual Recognition Challenge 2010A dataset of a billion masks plays the role ImageNet played for classification.
The 2010 paper on crowdworkers lists ImageNet among the work done on this marketplace: labelling millions of images became possible because labour could be ordered through an API.
Who are the Crowdworkers? Shifting Demographics in Mechanical TurkThe Inception v3 network was first trained on about 1.28 million images of 1,000 ImageNet classes from 2014 and only then fine-tuned on skin images.
Dermatologist-level classification of skin cancer with deep neural networksThe authors name ImageNet as the model: the dataset is meant to give audio event recognition what ImageNet gave object recognition in images.
Audio Set: An ontology and human-labeled dataset for audio eventsImageNet in 2009 compares itself with the ESP dataset: the game's labels sit at the "basic level", do not disambiguate word senses and are public for only 60,000 images.
ImageNetThe ImageNet paper calls LabelMe complementary to itself: about 200 categories and 30,000 images, but with the outlines and locations of objects.
ImageNetImageNet sets itself against this collection: here a synset averages 1000 images, 10-25% of them clean; in ImageNet about 99%, at full resolution.
ImageNet