gptkbp:instance_of
|
gptkb:Data
|
gptkbp:available_at
|
gptkb:Open_ML
gptkb:Kaggle
gptkb:Git_Hub
|
gptkbp:challenges
|
data quality
overfitting
class imbalance
computational efficiency
model generalization
|
gptkbp:contains
|
100,000 images
200 classes
|
gptkbp:created_by
|
gptkb:Stanford_University
|
gptkbp:developed_by
|
gptkb:Image_Net
|
gptkbp:downloadable_content
|
http://tiny-imagenet.herokuapp.com/
|
gptkbp:evaluates
|
top-5 accuracy
top-1 accuracy
|
https://www.w3.org/2000/01/rdf-schema#label
|
Tiny Image Net
|
gptkbp:image_format
|
gptkb:JPEG
|
gptkbp:image_labeling
|
gptkb:manual
|
gptkbp:is_divided_into
|
training, validation, test
|
gptkbp:is_notable_for
|
small size
hyperparameter tuning
model evaluation
ease of use
visualization techniques
feature extraction
performance benchmarking
research applications
data exploration
algorithm testing
educational value
diverse classes
|
gptkbp:is_part_of
|
gptkb:Image_Net_family
|
gptkbp:is_popular_in
|
deep learning community
|
gptkbp:is_similar_to
|
gptkb:CIFAR-10
gptkb:CIFAR-100
|
gptkbp:is_used_in
|
gptkb:academic_research
educational purposes
competitions
benchmarking algorithms
|
gptkbp:performance
|
machine learning models
|
gptkbp:provides_information_on
|
used
|
gptkbp:release_year
|
gptkb:2015
|
gptkbp:resolution
|
64x64 pixels
|
gptkbp:size
|
64x64 pixels
|
gptkbp:supports
|
transfer learning
convolutional neural networks
data science research
computer vision tasks
|
gptkbp:test_set_size
|
10,000 images
|
gptkbp:training_set_size
|
100,000 images
|
gptkbp:used_for
|
image classification
|
gptkbp:validation_set_size
|
10,000 images
|
gptkbp:bfsParent
|
gptkb:Wide_Res_Net
gptkb:Res_Net-101
gptkb:Efficient_Net_B3
|
gptkbp:bfsLayer
|
6
|