Gm Global Technolgoy Operations, Inc.

Andrea Forgacs

Ron M Hecht

Omer Tsimhoni

Dan Levi

Ariel Telpaz

Asaf Degani

Yael Shmueli Friedland

Michael Baltaxe

Yi G Glaser

Shaul Oron

Daniel S Glaser

Ronit Bustin

Gershon Celniker

Ohad Rahamim

A method of training a disparity estimation network. The method includes obtaining an eye-gaze dataset having first images with at least one gaze direction associated with each of the first images. A gaze prediction neural network is trained based on the eye-gaze dataset to develop a model trained to provide a gaze prediction for an external image. A depth database is obtained that includes second images having depth information associated with each of the second images. A disparity estimation neural network for object detection is trained based on an output from the gaze prediction neural network and an output from the depth database.

System and method for eye-gaze direction-based pre-training of neural networks

A system for vehicle trajectory planning. The system may include a polarized camera system configured to generate polarized images of a roadway to be driven over with a vehicle, a prior controller configured to generate a prior for the roadway based at least in part on the polarized images, and a driving assistance system configured to provide a driving assistance according to the prior.

Vehicle trajectory planning

A system and method for limited rate context-based eye gaze encoding includes generating, based on an outward looking camera situated in a vehicle, a first video stream of a surrounding environment to a controller. The controller generates a scene description based on the first video stream and selects a corresponding scene from a predefined list of known scenes. Further, the controller, based on the selected corresponding scene, selects a codebook of encoding and decoding parameters from a plurality of predefined codebooks. Based on an inward looking camera situated in the vehicle, a second video stream of a face of a driver to an eye tracker controller is generated, where the eye tracker controller estimates a gaze direction of the driver. Encoded data comprising the estimated gaze direction of the driver is transmitted over a bandwidth limited channel to a decoder, based on the selected codebook, the encoded data.

Limited rate context based eye gaze encoding

Growing community of inventors

Kfar Sava, Israel

Andrea Forgacs