Technical Description
The present invention relates to a composite backbone network-based system with enhanced and suppressed filter for 3D visual saliency that predicts exact and accurate salient objects in complex and cluttered backgrounds.
Problems Addressed
- Low Accuracy
- Limitations of Existing Technologies
- Challenges with Low-Depth Images
- Inaccurate Saliency Detection
- Poor Object Boundary Preservation
Tech Features
- High Detection Accuracy
- Multi-Modal Feature Integration
- Feature Representation & Preservation
- Deep Learning Framework
- Precise Salient Object Detection
- Self-Supervised Model
Target Audience
- Autonomous Vehicle Industry
- Smart Transportation Industry
- Aerospace & Defense Sector
- Research & Development
Tech ID: P02-1373 TRL 4 Patent Status: Published Available For Exclusive and Non-exclusive License
