Minmin Chen, Alice Zheng, and Kilian Q. Weinberger
Automatic image annotation is a diffifficult and highly relevant machine learning task. Recent advances have significantly improved the state-of-the-art in retrieval accuracy with algorithms based on nearest neighbor classification in carefully learned metric spaces. But this comes at a price of increased computational complexity during training and testing. We propose FastTag, a novel algorithm that achieves comparable results with two simple linear mappings that are co-regularized in a joint convex loss function. The loss function can be effifficiently optimized in closed form updates, which allows us to incorporate a large number of image descriptors cheaply. On several standard real-world benchmark data sets, we demonstrate that FastTag matches the current state-of-the-art in tagging quality, yet reduces the training and testing times by several orders of magnitude and has lower asymptotic complexity.
In Proceedings of the 30th International Conference on Machine Learning