Learning detectors quickly using structured covariance matrices
This work addresses the need for rapid detector estimation in computer vision, offering a significant speedup for applications requiring quick training, though it builds on prior insights about stationary negative distributions.
The paper tackles the problem of slow object detector training in computer vision by leveraging structured covariance matrices, achieving identical performance in orders of magnitude less time and memory compared to existing methods.
Computer vision is increasingly becoming interested in the rapid estimation of object detectors. Canonical hard negative mining strategies are slow as they require multiple passes of the large negative training set. Recent work has demonstrated that if the distribution of negative examples is assumed to be stationary, then Linear Discriminant Analysis (LDA) can learn comparable detectors without ever revisiting the negative set. Even with this insight, however, the time to learn a single object detector can still be on the order of tens of seconds on a modern desktop computer. This paper proposes to leverage the resulting structured covariance matrix to obtain detectors with identical performance in orders of magnitude less time and memory. We elucidate an important connection to the correlation filter literature, demonstrating that these can also be trained without ever revisiting the negative set.