18107173. TEMPORAL SEMANTIC BOUNDARY LOSS FOR VIDEO SEMANTIC SEGMENTATION NETWORKS simplified abstract (Samsung Electronics Co., Ltd.)
Contents
- 1 TEMPORAL SEMANTIC BOUNDARY LOSS FOR VIDEO SEMANTIC SEGMENTATION NETWORKS
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 TEMPORAL SEMANTIC BOUNDARY LOSS FOR VIDEO SEMANTIC SEGMENTATION NETWORKS - 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 Unanswered Questions
- 1.11 Original Abstract Submitted
TEMPORAL SEMANTIC BOUNDARY LOSS FOR VIDEO SEMANTIC SEGMENTATION NETWORKS
Organization Name
Inventor(s)
Mostafa El-khamy of San Diego CA (US)
Qingfeng Liu of San Diego CA (US)
TEMPORAL SEMANTIC BOUNDARY LOSS FOR VIDEO SEMANTIC SEGMENTATION NETWORKS - A simplified explanation of the abstract
This abstract first appeared for US patent application 18107173 titled 'TEMPORAL SEMANTIC BOUNDARY LOSS FOR VIDEO SEMANTIC SEGMENTATION NETWORKS
Simplified Explanation
The abstract describes a method for semantic segmentation in a network using input data from multiple frames to generate ground truth temporal semantic boundary maps and predicted temporal semantic boundary maps.
- Receiving input data from multiple frames in a semantic segmentation network
- Computing ground truth labels on the frames
- Generating ground truth temporal semantic boundary maps from the labels
- Generating predicted temporal semantic boundary maps based on the input data output
- Determining loss based on the ground truth and predicted temporal semantic boundary maps
Potential Applications
This technology could be applied in various fields such as autonomous driving, video surveillance, and medical imaging for accurate object detection and tracking.
Problems Solved
This technology solves the problem of accurately segmenting and tracking objects over time in video data, which is crucial for tasks like object recognition and scene understanding.
Benefits
The benefits of this technology include improved accuracy in semantic segmentation, better object tracking over time, and enhanced performance in various computer vision applications.
Potential Commercial Applications
Potential commercial applications of this technology include video analytics software for security systems, autonomous vehicle systems, and medical imaging software for diagnostic purposes.
Possible Prior Art
One possible prior art for this technology could be the use of recurrent neural networks for temporal segmentation in video data.
Unanswered Questions
How does this method compare to existing techniques in terms of accuracy and efficiency?
The article does not provide a comparison with existing techniques in terms of accuracy and efficiency.
What are the limitations of this method in real-world applications?
The article does not discuss the limitations of this method in real-world applications.
Original Abstract Submitted
Disclosed is a method including receiving, in a semantic segmentation network, input data from a plurality of frames, computing a ground truth label on the plurality of frames, generating a ground truth temporal semantic boundary map from the ground truth label on the plurality of frames, generating a predicted temporal semantic boundary map based on an output of the input data, and determining a loss based on the ground truth temporal semantic boundary map and the predicted temporal semantic boundary map.
