18153480. DEPTH ESTIMATION USING IMAGE AND SPARSE DEPTH INPUTS simplified abstract (QUALCOMM Incorporated)
Contents
- 1 DEPTH ESTIMATION USING IMAGE AND SPARSE DEPTH INPUTS
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 DEPTH ESTIMATION USING IMAGE AND SPARSE DEPTH INPUTS - A simplified explanation of the abstract
- 1.4 Simplified Explanation
- 1.5 Potential Applications
- 1.6 Problems Solved
- 1.7 Benefits
- 1.8 Potential Commercial Applications
- 1.9 Possible Prior Art
- 1.10 Original Abstract Submitted
DEPTH ESTIMATION USING IMAGE AND SPARSE DEPTH INPUTS
Organization Name
Inventor(s)
Georgi Dikov of Amsterdam (NL)
Mohsen Ghafoorian of Diemen (NL)
Joris Johannes Lambertus Van Vugt of Utrecht (NL)
DEPTH ESTIMATION USING IMAGE AND SPARSE DEPTH INPUTS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18153480 titled 'DEPTH ESTIMATION USING IMAGE AND SPARSE DEPTH INPUTS
Simplified Explanation
The abstract describes a method for generating depth information from images using a neural network model.
- The method involves obtaining an image of a scene and depth information associated with objects in the scene.
- An encoder of a neural network model processes the image and depth information to generate a feature representation.
- A decoder of the neural network model processes the feature representation to generate a depth output corresponding to the image.
Potential Applications
This technology can be applied in various fields such as:
- Augmented reality
- Autonomous driving
- Robotics
Problems Solved
This technology helps in:
- Enhancing the accuracy of depth perception in computer vision systems
- Improving object recognition and tracking in real-time applications
Benefits
The benefits of this technology include:
- Better understanding of 3D environments from 2D images
- Enhanced depth estimation for various applications
- Improved performance of depth-based algorithms
Potential Commercial Applications
This technology can be commercially utilized in:
- Camera systems for smartphones and tablets
- Surveillance systems
- Virtual reality headsets
Possible Prior Art
One possible prior art for this technology could be the use of traditional computer vision techniques for depth estimation from images.
What is the accuracy rate of depth estimation achieved by this technology?
The accuracy rate of depth estimation achieved by this technology can vary depending on the specific implementation and training data used.
How does this technology compare to LiDAR-based depth sensing systems?
This technology offers a cost-effective alternative to LiDAR-based depth sensing systems, providing depth information from images without the need for additional hardware.
Original Abstract Submitted
Systems and techniques are provided for generating depth information from one or more images. For instance, a method can include obtaining an image of a scene and obtaining depth information associated with one or more objects in the scene. The method can include processing, using an encoder of a neural network model, the image and the depth information to generate a feature representation of the image and the depth information. The method can further include processing, using a decoder of the neural network model, the feature representation of the image and the depth information to generate a depth output corresponding to the image.
