Are all the frames equally important?
This work addresses the need for temporal saliency datasets to train predictive algorithms in video analysis, offering a more accessible method for data collection, though it is incremental as it builds on existing saliency concepts.
The authors tackled the problem of measuring and predicting temporal video saliency, which assesses the importance of entire frames for human attention, by developing an interactive cursor-based algorithm to collect the first human response data. They showed that the scores qualitatively reflect semantic changes in frames and are highly correlated between observers, while the tool can collect fixations affordably without special equipment.
In this work, we address the problem of measuring and predicting temporal video saliency - a metric which defines the importance of a video frame for human attention. Unlike the conventional spatial saliency which defines the location of the salient regions within a frame (as it is done for still images), temporal saliency considers importance of a frame as a whole and may not exist apart from context. The proposed interface is an interactive cursor-based algorithm for collecting experimental data about temporal saliency. We collect the first human responses and perform their analysis. As a result, we show that qualitatively, the produced scores have very explicit meaning of the semantic changes in a frame, while quantitatively being highly correlated between all the observers. Apart from that, we show that the proposed tool can simultaneously collect fixations similar to the ones produced by eye-tracker in a more affordable way. Further, this approach may be used for creation of first temporal saliency datasets which will allow training computational predictive algorithms. The proposed interface does not rely on any special equipment, which allows to run it remotely and cover a wide audience.