ABSTRACT
This paper introduces AdaROI-Depth, a method that uses depth information to adaptively select regions of interest for enhancing event frame reconstruction. It integrates depth-guided attention with adaptive pooling to identify important regions for reconstruction, improving quality while maintaining efficiency.
PAPER · PDF
Loading PDF…
Key findings
AdaROI-Depth leverages depth information to enhance event frame reconstruction.
The method integrates depth-guided attention with adaptive pooling.
Significant improvement in reconstruction quality is demonstrated.
The approach maintains computational efficiency.
Limitations & open questions
Further research is needed to scale the method for real-time applications.