Inventor

GAIDON ADRIEN

US28 patents

⚠️ This page may combine multiple inventors who share the name “GAIDON ADRIEN”. Patents are grouped by organization below to help distinguish them — per-person disambiguation is on the roadmap.

GAIDON ADRIEN

TOYOTA RES INST INC

Systems and methods for self-supervised depth estimation

Spatiotemporal relationship reasoning for pedestrian intent prediction

Systems and methods for multi-camera modeling with neural camera networks

Disentangling human dynamics for pedestrian locomotion forecasting with noisy supervision

Systems and methods for generic visual odometry using learned features via neural camera models

Self-calibration for decoration based sensor fusion method

Systems and methods for multi-camera modeling with neural camera networks

Systems and methods for end-to-end map building from a video sequence using neural camera models

Systems and methods for generic visual odometry using learned features via neural camera models

Variational 3D object detection

Self-supervised depth for volumetric rendering regularization

Cross-attention decoding for volumetric rendering

Cooperative-contrastive learning systems and methods

Attention-based recurrent convolutional network for vehicle taillight recognition

System and method for monocular depth estimation from semantic information

Self extrinsic self-calibration via geometrically consistent self-supervised depth and ego-motion learning

Guided domain randomization via differentiable dataset rendering

Systems and methods for neural implicit scene representation with dense, uncertainty-aware monocular depth constraints

Reinforcement learning based control of imitative policies for autonomous driving

XEROX CORP

Multi-object tracking with generic object proposals

Extracting gradient features from neural networks

Self-learning object detectors for unlabeled videos using multi-task learning

Deep data association for online multi-class multi-object tracking

Efficient object detection with patch-level window processing

Generating a virtual world to assess real-world video analysis performance

Online domain adaptation for multi-object tracking

System and method for video classification using a hybrid unsupervised and supervised multi-layer architecture

TOYOTA RES INSTITUTE

Method and system for learning reward functions for driving using positive-unlabeled reward learning