Saliency
Shallow and deep convolutional neural networks for predicting visual saliency in images using a data-driven approach.
What is Saliency?
Saliency-2016-cvpr is a research project that provides two convolutional neural network models for predicting visual saliency in images. It solves the problem of identifying which parts of an image attract human attention by using a purely data-driven, end-to-end learning approach instead of relying on hand-crafted features. The project includes both a shallow network trained from scratch and a deeper network with transferred features from classification tasks.
Computer vision researchers and practitioners working on visual attention modeling, saliency prediction, or applications requiring understanding of image regions that capture human gaze. It's particularly relevant for those interested in early deep learning approaches to this problem.
Developers choose this project because it provides well-documented, pre-trained models from a peer-reviewed CVPR publication, offering both lightweight and deeper architectural options. It represents one of the first end-to-end CNN implementations specifically for saliency prediction, serving as a foundational reference in the field.
Overview
Shallow and Deep Convolutional Networks for Saliency Prediction
Use Cases
Best For
Related Projects
Found a gem we're missing?
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
