'Which imagenet subset normally evaluated on in image classification thesis?

Speaking of imagenet, I found that the ground truth of val subset is available, the test subset is not however.

Thus, I'm curious about the classification evaluation in thesis. The accuracy is normally evaluated on test or val set?

conv-neural-networkclassificationimage-classificationimagenet


Sources

This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.

Source: Stack Overflow

Solution Source

Related Questions

Sparse Categorical CrossEntropy shape problem with Keras

Metrics F1 warning zero division

Weight tensor should be defined either for all 1000 classes or no classes but got weight tensor of shape: [5]

Scikit-Learn: How to retrieve prediction probabilities for a KFold CV?

Which algorithm is the best based on P-Values

Only two distinct values of probabilities in Logistic regression output

Difference between absdiff and normal subtraction in OpenCV

ValueError: Input 0 of layer is incompatible with the layer: expected shape=(None, 224, 224, 3), found shape=(224, 224, 3)

Classification metrics can't handle a mix of binary and continuous targets [duplicate]

Scalable/Iterative Large Data Frame Dimensionality Reduction R

SKLearn: Getting distance of each point from decision boundary?

tensorflow automatic accuracy calculation for multilabel classifier

Neural Network for Regression with tflearn

Multi-output neural network combining regression and classification

Siamese Network for binary classification with pre-encoded inputs