18056386. INFORMATION PROCESSING SYSTEM AND INFORMATION PROCESSING METHOD simplified abstract (CANON KABUSHIKI KAISHA)
Contents
INFORMATION PROCESSING SYSTEM AND INFORMATION PROCESSING METHOD
Organization Name
Inventor(s)
Ritsuya Tomita of Kanagawa (JP)
INFORMATION PROCESSING SYSTEM AND INFORMATION PROCESSING METHOD - A simplified explanation of the abstract
This abstract first appeared for US patent application 18056386 titled 'INFORMATION PROCESSING SYSTEM AND INFORMATION PROCESSING METHOD
Simplified Explanation
An information processing system is described in this patent application. Here are the key points:
- The system includes a first information processing apparatus and a second information processing apparatus.
- The first apparatus has a training data acquisition unit to acquire training data.
- It also has a first learning unit that performs first learning processing on a first partial model, which includes the input layer and a part of the intermediate layers of an inference model.
- The second apparatus has a second learning unit that performs second learning processing on a second partial model, which includes an intermediate layer different from the first partial model.
- The third learning unit in the first apparatus performs third learning processing on a third partial model, which includes the output layer, using the output obtained from the second learning processing and the correct label.
Potential applications of this technology:
- This information processing system can be used in various machine learning tasks, such as image recognition, natural language processing, and speech recognition.
- It can be applied in industries like healthcare, finance, manufacturing, and transportation to improve data analysis and decision-making processes.
Problems solved by this technology:
- The system allows for distributed learning, where different parts of the model are trained on different apparatuses, enabling faster and more efficient training.
- It addresses the challenge of training large-scale models by dividing the learning process into multiple partial models and distributing the workload across multiple apparatuses.
Benefits of this technology:
- By distributing the learning process, the system reduces the training time required for complex models.
- It improves the scalability of the learning process, allowing for larger models to be trained.
- The system enables collaborative learning, where multiple apparatuses can contribute to the training process, leading to improved accuracy and performance of the model.
Original Abstract Submitted
An information processing system includes a first information processing apparatus including a training data acquisition unit configured to acquire training data, a first learning unit configured to perform first learning processing by inputting the learning data to a first partial model including the input layer and a part of a plurality of intermediate layers of an inference model, and a third learning unit configured to perform third learning processing on a third partial model including the output layer using an output obtained through second learning processing performed by a second information processing apparatus and the correct label, and the second information processing apparatus including a second learning unit configured to perform the second learning processing by inputting an output obtained through the first learning processing to a second partial model including an intermediate layer different from the part of the plurality of intermediate layers included in the first partial model.
