US 12,243,300 B2
Efficient object detection using deep learning techniques
Soyeb Nagori, Bangalore (IN); and Deepak Poddar, Bangalore (IN)
Assigned to Texas Instruments Incorporated, Dallas, TX (US)
Filed by TEXAS INSTRUMENTS INCORPORATED, Dallas, TX (US)
Filed on Oct. 27, 2021, as Appl. No. 17/512,049.
Claims priority of application No. 202041049581 (IN), filed on Nov. 12, 2020.
Prior Publication US 2022/0147748 A1, May 12, 2022
Int. Cl. G06V 10/25 (2022.01); G06N 3/048 (2023.01); G06V 10/94 (2022.01)
CPC G06V 10/95 (2022.01) [G06N 3/048 (2023.01); G06V 10/25 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A method, comprising:
determining, using a device based on a machine learning model, for each location of a set of locations in an image, a set of confidence levels corresponding to a set of objects, wherein the set of confidence levels each represents a probability of presence of a respective one of the set of objects at the location;
determining, using the device, for each location of the set of locations, an upper-bound score based on the set of confidence levels at the location;
determining, using the device, a subset of the set of locations based on the upper-bound scores of the set of locations and a confidence threshold; and
detecting, using the device, the set of objects in the image based on the subset of locations.