Inventor · Albany, CA, US

Kratarth Goel

20Patents
3h-index
19Co-inventors
52Inventor score

Filing activity: Jun 20, 2018 → Jun 5, 2023

Most-cited inventions

PatentTitleAreaCited byStatus
US10984290B1 Multi-task learning for real-time semantic and/or depth aware instance segmentation and/or three-dimensional object bounding Physics 10 Active
US11710352B1 Machine-learned model training for pedestrian attribute and gesture detection Physics 6 Active
US10817740B2 Instance segmentation inferred from machine learning model output Physics 5 Active
US11610078B2 Low variance region detection for improved high variance region detection using machine learning Physics 3 Active
US10936922B2 Machine learning techniques Physics 2 Active
US11592818B2 Restricted multi-scale inference for machine learning Physics 2 Active
US11450117B2 Hierarchical machine-learning network architecture Physics 2 Active
US12051276B1 Machine-learned model training for pedestrian attribute and gesture detection Physics 1 Active
US11893750B2 Multi-task learning for real-time semantic and/or depth aware instance segmentation and/or three-dimensional object bounding Physics 1 Active
US11605236B2 Training a machine-learned model to detect low variance regions Performing Operations; Transporting 1 Active
US12387500B1 Parallel vehicle model processing Performing Operations; Transporting 0 Active
US12361567B1 Multi-frame temporal aggregation and dense motion estimation for autonomous vehicles Physics 0 Active
US12236705B1 Pedestrian attribute and gesture detection Physics 0 Active
US12100224B1 Key point detection Physics 0 Active
US11163990B2 Vehicle control system and method for pedestrian detection based on head detection in sensor data Physics 0 Active
US11681046B2 Depth data model training with upsampling, losses and loss balancing Physics 0 Active
US12084060B2 Associating image data and depth data for depth determination Physics 0 Active
US12056934B2 Three-dimensional object detection based on image data Physics 0 Active
US10963709B2 Hierarchical machine-learning network architecture Physics 0 Active
US11157774B2 Depth data model training with upsampling, losses, and loss balancing Physics 0 Active

Source: USPTO / EPO open patent data. Inventor disambiguation is heuristic; counts are objective bibliographic measures.