Faster gaze prediction with dense networks and Fisher pruning
This work addresses the need for faster gaze prediction, which is crucial for real-time applications and video saliency models, though it is incremental as it builds on existing deep learning methods.
The paper tackles the problem of overparameterization in deep networks for gaze prediction by introducing Fisher pruning combined with knowledge distillation, achieving a 10x speedup while maintaining state-of-the-art AUC performance on the CAT2000 dataset.
Predicting human fixations from images has recently seen large improvements by leveraging deep representations which were pretrained for object recognition. However, as we show in this paper, these networks are highly overparameterized for the task of fixation prediction. We first present a simple yet principled greedy pruning method which we call Fisher pruning. Through a combination of knowledge distillation and Fisher pruning, we obtain much more runtime-efficient architectures for saliency prediction, achieving a 10x speedup for the same AUC performance as a state of the art network on the CAT2000 dataset. Speeding up single-image gaze prediction is important for many real-world applications, but it is also a crucial step in the development of video saliency models, where the amount of data to be processed is substantially larger.