Inventor
SUN DEQING
US16 patents
⚠️ This page may combine multiple inventors who share the name “SUN DEQING”. Patents are grouped by organization below to help distinguish them — per-person disambiguation is on the roadmap.
NVIDIA CORP
13 patentsUS11636668B2Apr 25, 2023
Bilateral convolution layer network for processing point clouds
NVIDIA CORP7 citations85
US10424069B2Sep 24, 2019
System and method for optical flow estimation
NVIDIA CORP8 citations84
US10776688B2Sep 15, 2020
Multi-frame video interpolation using optical flow
NVIDIA CORP11 citations83
US10748036B2Aug 18, 2020
Training a neural network to predict superpixels using segmentation-aware affinity loss
NVIDIA CORP7 citations82
US11082720B2Aug 3, 2021
Using residual video data resulting from a compression of original video data to improve a decompression of the original video data
NVIDIA CORP2 citations73
US10467763B1Nov 5, 2019
System and method for optical flow estimation
NVIDIA CORP3 citations73
US10929987B2Feb 23, 2021
Learning rigidity of dynamic scenes for three-dimensional scene flow estimation
NVIDIA CORP2 citations71
US10986325B2Apr 20, 2021
Scene flow estimation using shared features
NVIDIA CORP4 citations67
US11496773B2Nov 8, 2022
Using residual video data resulting from a compression of original video data to improve a decompression of the original video data
NVIDIA CORP0 citations62
US11508076B2Nov 22, 2022
Learning rigidity of dynamic scenes for three-dimensional scene flow estimation
NVIDIA CORP0 citations60
US11256961B2Feb 22, 2022
Training a neural network to predict superpixels using segmentation-aware affinity loss
NVIDIA CORP0 citations60
US12566963B1Mar 3, 2026
Convolutional neural networks with content-adaptive filters
NVIDIA CORP0 citations59
US10789678B2Sep 29, 2020
Superpixel sampling networks
NVIDIA CORP0 citations52
GOOGLE LLC
3 patentsUS11790550B2Oct 17, 2023
Learnable cost volume for determining pixel correspondence
GOOGLE LLC2 citations65
US12573161B2Mar 10, 2026
Learning articulated shape reconstruction from imagery
GOOGLE LLC0 citations47
US11954899B2Apr 9, 2024
Systems and methods for training models to predict dense correspondences in images using geodesic distances
GOOGLE LLC0 citations47