Top-k Multiclass SVM
This work addresses image classification challenges with many classes by allowing k guesses, which is incremental as it builds on the multiclass SVM framework.
The authors tackled the problem of class ambiguity in image classification by proposing a top-k multiclass SVM that directly optimizes for top-k error, resulting in consistent improvements in top-k accuracy across five datasets.
Class ambiguity is typical in image classification problems with a large number of classes. When classes are difficult to discriminate, it makes sense to allow k guesses and evaluate classifiers based on the top-k error instead of the standard zero-one loss. We propose top-k multiclass SVM as a direct method to optimize for top-k performance. Our generalization of the well-known multiclass SVM is based on a tight convex upper bound of the top-k error. We propose a fast optimization scheme based on an efficient projection onto the top-k simplex, which is of its own interest. Experiments on five datasets show consistent improvements in top-k accuracy compared to various baselines.