Places205-VGG, a CNN trained model for 205 scene categories of Places Database with 2.5 million images Places205 dataset has top1 accuracy = 58.9% and top5 accuracy = 87.7%.
What does top1 and top5 (and, in general, top $N$) accuracy mean in the context of deep learning?
It is explained in this CrossValidated post.
Top1 accuracy means the best guess (class with highest probability) is the correct result 58.9% of the time, while top5 accuracy means the correct result is in the top 5 best guesses (5 classes with highest probabilities) 87.7% of the time.
