JVCKENWOOD had 20 patents in artificial intelligence during Q4 2023. JVCKENWOOD Corp’s patents in Q4 2023 focus on improving machine learning processes. This includes determining initialization rates for neural network models based on layer depth, controlling detection functions for dangerous driving based on vehicle situation, adapting models through transfer learning with domain adaptation data richness, optimizing filter training in continual learning, and generating virtual image data for machine learning using principal component analysis. GlobalData’s report on JVCKENWOOD gives a 360-degreee view of the company including its patenting strategy. Buy the report here.
Smarter leaders trust GlobalData
Discover B2B Marketing That Performs
Combine business intelligence and editorial excellence to reach engaged professionals across 36 leading media platforms.
JVCKENWOOD grant share with artificial intelligence as a theme is 10% in Q4 2023. Grant share is based on the ratio of number of grants to total number of patents.
Recent Patents
Application: Machine learning device, inference device, and machine learning method (Patent ID: US20230409912A1)
The patent filed by JVCKENWOOD Corp. describes a machine learning device and method that involves determining an initialization rate for weights in a neural network model based on the depth of a layer in the model. The device includes units for initialization rate determination, machine learning execution, and initialization, which work together to generate a neural network model trained on a first task and then initialize the weights in the model for use in a second task. The initialization rate determination unit adjusts the initialization rate based on the layer's depth, with smaller rates for layers closer to the input layer and larger rates for layers closer to the output layer.
Additionally, the device can determine a second initialization rate based on the similarity between the first and second tasks, allowing for transfer learning from the initialized neural network trained on the first task to generate a model for the second task. The larger the similarity between tasks, the larger the second initialization rate. Furthermore, the patent also describes an inference device that selects tasks, generates an inference neural network model, and infers the selected task based on the model. Overall, the method and devices outlined in the patent aim to optimize the training and initialization process for neural network models across multiple tasks efficiently.
To know more about GlobalData’s detailed insights on JVCKENWOOD, buy the report here.
Data Insights
From
The gold standard of business intelligence.
Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.
GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.
GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.
