CIFAR-100
E367299
CIFAR-100 is a widely used image classification dataset consisting of 60,000 32×32 color images across 100 object categories, commonly used to benchmark machine learning models.
All labels observed (4)
| Label | Occurrences |
|---|---|
| CIFAR-100 canonical | 4 |
| CIFAR-100 dataset | 1 |
| CIFAR-FS | 1 |
| CIFAR100 | 1 |
How this entity was disambiguated
This entity first appeared as the object of triple T3542985 — resolving that mention is where its identity was fixed. The disambiguator weighed these candidate entities and picked the highlighted one (or “None”, minting a new entity). This is how homonymy is resolved: the same surface form can point to different entities.
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: CIFAR-100 Context triple: [ResNet, benchmarkedOn, CIFAR-100]
-
A.
ImageNet
ImageNet is a large-scale visual database widely used for training and benchmarking image classification and computer vision algorithms.
-
B.
CIFAR
CIFAR (the Canadian Institute for Advanced Research) is a Canadian global research organization that supports long-term, collaborative, interdisciplinary research, including major initiatives in artificial intelligence.
-
C.
MNIST
MNIST is a widely used benchmark dataset of handwritten digit images commonly employed for training and evaluating image classification algorithms in machine learning and computer vision.
-
D.
Fashion-MNIST
Fashion-MNIST is a popular benchmark dataset of Zalando clothing item images used as a more challenging drop-in replacement for the original MNIST handwritten digits in machine learning research.
-
E.
KMNIST
KMNIST is a benchmark image dataset of handwritten Japanese characters (hiragana) designed as a more complex, drop-in replacement for the original MNIST digit dataset.
- F. None of above. chosen
- G. Unsure - the case is ambiguous/there is not enough information to decide.
NED2
Entity disambiguation (via description)
gpt-5-mini-2025-08-07
Target entity: CIFAR-100 Target entity description: CIFAR-100 is a widely used image classification dataset consisting of 60,000 32×32 color images across 100 object categories, commonly used to benchmark machine learning models.
-
A.
ImageNet
ImageNet is a large-scale visual database widely used for training and benchmarking image classification and computer vision algorithms.
-
B.
CIFAR
CIFAR (the Canadian Institute for Advanced Research) is a Canadian global research organization that supports long-term, collaborative, interdisciplinary research, including major initiatives in artificial intelligence.
-
C.
MNIST
MNIST is a widely used benchmark dataset of handwritten digit images commonly employed for training and evaluating image classification algorithms in machine learning and computer vision.
-
D.
Fashion-MNIST
Fashion-MNIST is a popular benchmark dataset of Zalando clothing item images used as a more challenging drop-in replacement for the original MNIST handwritten digits in machine learning research.
-
E.
KMNIST
KMNIST is a benchmark image dataset of handwritten Japanese characters (hiragana) designed as a more complex, drop-in replacement for the original MNIST digit dataset.
- F. None of above. chosen
Statements (48)
| Predicate | Object |
|---|---|
| instanceOf |
benchmark dataset
ⓘ
computer vision dataset ⓘ image classification dataset ⓘ |
| accessMethod | download from official website ⓘ |
| commonlyUsedWith | CIFAR-10 ⓘ |
| developedAt | University of Toronto ⓘ |
| developedBy |
Alex Krizhevsky
ⓘ
Geoffrey Hinton ⓘ |
| difficultyLevel | harder than CIFAR-10 ⓘ |
| domain | natural images ⓘ |
| evaluationMetric |
classification accuracy
ⓘ
top-1 accuracy ⓘ top-5 accuracy ⓘ |
| fileFormat | binary batch files ⓘ |
| hasCanonicalURL | https://www.cs.toronto.edu/~kriz/cifar.html ⓘ |
| hasCoarseGrainedSuperclasses | 20 ⓘ |
| hasColorChannels | 3 ⓘ |
| hasColorSpace | RGB ⓘ |
| hasFineGrainedClasses | 100 ⓘ |
| hasImageHeight | 32 ⓘ |
| hasImageResolution | 32x32 pixels ⓘ |
| hasImageWidth | 32 ⓘ |
| hasInputType | 32x32 RGB image tensor ⓘ |
| hasLabelStructure |
coarse labels
ⓘ
fine labels ⓘ |
| hasNumberOfClasses | 100 ⓘ |
| hasTestImages | 10000 ⓘ |
| hasTotalImages | 60000 ⓘ |
| hasTrainingImages | 50000 ⓘ |
| includedIn |
TensorFlow Datasets
ⓘ
TorchVision datasets ⓘ
linked to:
torchvision (ecosystem)
common deep learning libraries ⓘ |
| introducedAs | subset of the Tiny Images dataset ⓘ |
| language | label names in English ⓘ |
| license | MIT-like license ⓘ |
| partOf |
Canadian Institute for Advanced Research (CIFAR)
ⓘ
linked to:
Canadian Institute for Advanced Research
|
| taskType | multi-class classification ⓘ |
| timePeriod | late 2000s ⓘ |
| typicalSplit |
100 test images per class
ⓘ
500 training images per class ⓘ |
| usedFor |
architecture comparison
ⓘ
benchmarking deep learning models ⓘ benchmarking machine learning models ⓘ data augmentation research ⓘ image recognition research ⓘ regularization method evaluation ⓘ representation learning ⓘ transfer learning evaluation ⓘ |
How these facts were elicited
The pipeline generated the facts above by prompting gpt-5.1 with this entity's name + description and the instruction below.
Instruction
You are a knowledge base construction expert. Given a subject entity and a description of it, return factual statements that you know for the subject as a JSON list of dictionaries(triples), where keys must be "subject", "predicate" and "object". The number of facts may be very high, between 25 to 50 or more, for very popular subjects. For less popular subjects, the number of facts can be very low, like 5 or 10. # Requirements - If you don't know the subject at all, return an empty list. - If the subject is not a named entity, return an empty list. - Include at least one triple where predicate is "instanceOf". - Do not get too wordy. - Separate several objects into multiple triples with one object.
Input
Subject: CIFAR-100 Description of subject: CIFAR-100 is a widely used image classification dataset consisting of 60,000 32×32 color images across 100 object categories, commonly used to benchmark machine learning models.
Referenced by (7)
Full triples — surface form annotated when it differs from this entity's canonical label.
linked to: CIFAR-100
linked to: CIFAR-100