Toyota jidosha kabushiki kaisha (20240119354). MODEL TRAINING METHOD AND MODEL TRAINING SYSTEM simplified abstract
Contents
- 1 MODEL TRAINING METHOD AND MODEL TRAINING SYSTEM
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 MODEL TRAINING METHOD AND MODEL TRAINING SYSTEM - 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
MODEL TRAINING METHOD AND MODEL TRAINING SYSTEM
Organization Name
toyota jidosha kabushiki kaisha
Inventor(s)
Hsuan-Kung Yang of Minato-ku Tokyo (JP)
Norimasa Kobori of Nerima-ku Tokyo (JP)
MODEL TRAINING METHOD AND MODEL TRAINING SYSTEM - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240119354 titled 'MODEL TRAINING METHOD AND MODEL TRAINING SYSTEM
Simplified Explanation
The model training method described in the abstract is focused on training an object identification model using machine learning techniques. The method involves acquiring labeled training data where a track is given as a label to a sequence of images, representing the movement of the same object. The track is automatically obtained by a tracker that follows the object in the sequence of images. The object identification model is then trained based on this labeled training data.
- Acquiring labeled training data with tracks representing the movement of the same object in a sequence of images.
- Automatically obtaining tracks using a tracker that follows the object in the images.
- Training the object identification model based on the labeled training data.
Potential Applications
This technology can be applied in various fields such as surveillance, autonomous vehicles, robotics, and augmented reality for accurate object identification and tracking.
Problems Solved
1. Accurate object identification and tracking in sequences of images. 2. Automating the process of obtaining tracks for training data.
Benefits
1. Improved accuracy in object identification. 2. Efficiency in training object identification models. 3. Automation of the track acquisition process.
Potential Commercial Applications
Optimizing surveillance systems for better object tracking and identification.
Possible Prior Art
One possible prior art could be the use of manual labeling for training object identification models, which can be time-consuming and less accurate compared to the automated track acquisition method described in this patent application.
Unanswered Questions
How does the model handle occlusions or changes in appearance of the object during tracking?
The abstract does not provide information on how the model deals with occlusions or changes in appearance of the object during tracking.
What is the computational cost associated with training the object identification model using this method?
The abstract does not mention the computational cost involved in training the object identification model using the described method.
Original Abstract Submitted
a model training method trains an object identification model that is based on machine learning. the model training method includes acquiring labeled training data where a track is given as a label to a sequence of images. the track is information representing a time series of a same moving object in the sequence of images and is automatically obtained by a tracker that tracks the same moving object in the sequence of images. the model training method further includes training the object identification model based on the labeled training data.
