An Analysis of Object Representations in Deep Visual Trackers
This addresses a fundamental limitation in single object visual tracking for computer vision applications, but is incremental as it builds on existing deep correlation networks.
The paper tackled the problem that deep visual trackers often default to saliency detection instead of using object instance representations, and introduced an auxiliary detection task that improved tracking performance.
Fully convolutional deep correlation networks are integral components of state-of the-art approaches to single object visual tracking. It is commonly assumed that these networks perform tracking by detection by matching features of the object instance with features of the entire frame. Strong architectural priors and conditioning on the object representation is thought to encourage this tracking strategy. Despite these strong priors, we show that deep trackers often default to tracking by saliency detection - without relying on the object instance representation. Our analysis shows that despite being a useful prior, salience detection can prevent the emergence of more robust tracking strategies in deep networks. This leads us to introduce an auxiliary detection task that encourages more discriminative object representations that improve tracking performance.