Influence of Image Classification Accuracy on Saliency Map Estimation
This work addresses the problem of improving saliency map estimation for computer vision researchers by leveraging image classification models, though it is incremental as it builds on existing pretrained models.
The paper investigates the relationship between image classification accuracy and saliency map estimation performance, finding a strong correlation, and proposes a model with multi-scale images and upsampling layers that achieves state-of-the-art accuracy on PASCAL-S, OSIE, and MIT1003 datasets.
Saliency map estimation in computer vision aims to estimate the locations where people gaze in images. Since people tend to look at objects in images, the parameters of the model pretrained on ImageNet for image classification are useful for the saliency map estimation. However, there is no research on the relationship between the image classification accuracy and the performance of the saliency map estimation. In this paper, it is shown that there is a strong correlation between image classification accuracy and saliency map estimation accuracy. We also investigated the effective architecture based on multi scale images and the upsampling layers to refine the saliency-map resolution. Our model achieved the state-of-the-art accuracy on the PASCAL-S, OSIE, and MIT1003 datasets. In the MIT Saliency Benchmark, our model achieved the best performance in some metrics and competitive results in the other metrics.