Verisk Analytics. has been granted a patent for a computer vision system that utilizes reinforcement learning for object detection. The system identifies target objects in images, determines bounding boxes, and enhances detection accuracy through reinforcement learning based on received rewards. GlobalData’s report on Verisk Analytics gives a 360-degree view of the company including its patenting strategy. Buy the report here.

According to GlobalData’s company profile on Verisk Analytics, was a key innovation area identified from patents. Verisk Analytics's grant share as of July 2024 was 29%. Grant share is based on the ratio of number of grants to total number of patents.

Object detection using reinforcement learning in computer vision

Source: United States Patent and Trademark Office (USPTO). Credit: Verisk Analytics Inc

The granted patent US12067644B2 outlines a computer vision system and method for object detection utilizing reinforcement learning. The system comprises a memory that stores images and a processor that interacts with this memory. Key functionalities include setting reinforcement learning parameters, detecting target objects within images, and determining bounding boxes around these objects. The system enhances the detection process by performing reinforcement learning on the image portions within the bounding boxes, allowing for adjustments based on positive feedback received by the learning agent. The claims also detail various configurations for the bounding boxes, including options for magnification, centering, and splitting when multiple objects are detected.

Additionally, the patent describes methods for refining the bounding box based on the detected objects and their characteristics. The processor can select specific aspect ratios and sizes for the bounding boxes and can oppose movements in response to triggering actions. The system is designed to be trained on both single and multiple class object categories, enabling it to adaptively learn and initiate actions based on the presence of various objects in the images. Overall, the patent presents a comprehensive approach to improving object detection in computer vision through reinforcement learning techniques.

To know more about GlobalData’s detailed insights on Verisk Analytics, 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

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.