Ambarella. has been granted a patent for an apparatus that utilizes a processor and memory to detect objects in images. The system employs a quantized multi-stage object detection network, involving data range analyses, region proposal networks, and region-based convolutional neural networks to enhance object detection accuracy. GlobalData’s report on Ambarella gives a 360-degree view of the company including its patenting strategy. Buy the report here.

Access deeper industry intelligence

Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.

Find out more

According to GlobalData’s company profile on Ambarella, Cloud gaming was a key innovation area identified from patents. Ambarella's grant share as of July 2024 was 90%. Grant share is based on the ratio of number of grants to total number of patents.

Object detection using a quantized multi-stage network

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

The patent US12073599B2 describes an advanced apparatus and method for object detection utilizing a quantized multi-stage object detection network. The apparatus includes an interface for receiving image data and a processor that employs a series of steps to detect objects within the image. Key processes involve generating quantized image data through a first data range analysis, creating a feature map and proposal bounding boxes via a region proposal network (RPN), and selecting ground truth bounding boxes. The processor further generates region of interest pooling results, applies a second data range analysis, and utilizes a region-based convolutional neural network (RCNN) to refine the detection process. The claims also detail the training of the RPN and RCNN, the storage of these networks as directed acyclic graphs, and the sharing of convolution layers between them.

Additionally, the patent outlines specific techniques for enhancing the object detection process, such as cropping and resampling feature maps, performing bilinear interpolation for resizing, and utilizing a three-dimensional feature map structure. The apparatus is applicable in various fields, including computer vision systems and autonomous vehicles. The method claims mirror the apparatus claims, emphasizing the systematic approach to object detection through quantization and multi-stage processing, ensuring efficient and accurate identification of objects in image data.

To know more about GlobalData’s detailed insights on Ambarella, 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.