A Survey on Neural Radiance Fields

View Full Paper

Ask AI

Bookmark

Authors

Discussion

This survey reviews advancements in neural radiance fields for view synthesis, highlighting model comparisons and future challenges.

Neural radiance fields have revolutionized view synthesis in computer vision, enhancing previous techniques.
The survey analyzes seven NeRF-based forms, including implicit representation and neural point clouds.
Comparisons of 14 research directions reveal improvements in modeling dynamic scenes and generalization.
Qualitative and quantitative evaluations on multiple datasets provide insights into training time and memory needs.

Liao et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark