Summary

Introduction

• Introducing recent research on human attention modeling for enabling delightful user experiences.
• Highlighting two papers published in CVPR 2022 and CVPR 2023 related to this area.
• Discussion on the applications of predictive models of human attention.

Attention-guided image editing

• Explaining the optimization framework developed for guiding visual attention in images.
• Introducing the user-aware saliency model that can predict attention for one user, a group of users, and the general population with a single model.
• Describing the use of per-user attention maps and adaptive user masks in the model to encode individuals' attention patterns.
• Presenting attention predictions for two different groups of participants on two images.

Progressive image decoding centered on salient features

• Explaining the concept of progressive decoding of images for improving user experience in browsing.
• Describing how predictive attention models can help with image compression and faster loading of web pages with images.
• Highlighting the potential of predictive attention models for improving the robustness of computer vision models in object classification or detection tasks.

Conclusion

• Summarizing the benefits of predictive models of human attention for enabling delightful user experiences in various applications.