Nvidia corporation (20240282118). OBJECT DETECTION USING POLYGONS FOR AUTONOMOUS SYSTEMS AND APPLICATIONS simplified abstract
Contents
OBJECT DETECTION USING POLYGONS FOR AUTONOMOUS SYSTEMS AND APPLICATIONS
Organization Name
Inventor(s)
Yang Zheng of Thousand Oak CA (US)
Trung Pham of Santa Clara CA (US)
Minwoo Park of Saratoga CA (US)
OBJECT DETECTION USING POLYGONS FOR AUTONOMOUS SYSTEMS AND APPLICATIONS - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240282118 titled 'OBJECT DETECTION USING POLYGONS FOR AUTONOMOUS SYSTEMS AND APPLICATIONS
- Simplified Explanation:**
The patent application discusses how object detectors in systems like autonomous driving can regress bounding polygons for detected objects. This involves determining regression data for the shape of an object and generating a bounding shape based on this data.
- Key Features and Innovation:**
- Object detectors regress bounding polygons for detected objects.
- Regression data includes angles and lengths of vectors for the polygon.
- Bounding shapes are generated based on regression data.
- Object detectors can be trained by deforming polygons to match ground truth.
- Potential Applications:**
This technology can be applied in autonomous driving systems, surveillance systems, robotics, and any other system requiring object awareness, identification, avoidance, and localization.
- Problems Solved:**
This technology addresses the need for accurate object detection and localization in various systems, improving safety and efficiency.
- Benefits:**
- Enhanced object detection accuracy.
- Improved object localization.
- Increased system efficiency and safety.
- Commercial Applications:**
Title: Advanced Object Detection and Localization Technology for Autonomous Systems This technology can be utilized in autonomous vehicles, security systems, industrial automation, and smart city infrastructure, leading to safer and more efficient operations.
- Prior Art:**
Researchers can explore prior art related to object detection, regression algorithms, and geometric modeling in computer vision and machine learning literature.
- Frequently Updated Research:**
Researchers are continually developing new algorithms and techniques to improve object detection and localization accuracy in various applications.
- Questions about Object Detection and Localization Technology:**
1. How does this technology improve the accuracy of object detection in autonomous systems? 2. What are the key challenges in training object detectors to regress bounding polygons accurately?
Original Abstract Submitted
in various examples, one or more object detectors may regress bounding polygons for detected objects in systems (e.g., autonomous or semi-autonomous driving systems and applications) that provide object awareness, object identification, object avoidance, and/or object localization. the object detector may determine regression data representing a regressed polygon associated with a given shape of a detected object represented by classification data determined from a scene. the object detector may determine regression data for different regressed angles between different pairs of successive vertices of the regressed polygon and regressed lengths of vectors from a regressed geometric center of the regressed polygon to vertices of the regressed polygon. the object detector may generate, based at least in part on the regression data, a bounding shape for a detected object in the scene. in some embodiments, the object detector may be trained by deforming a regressed polygon to match a ground truth polygon.
