Z
Zhengping Ji
Researcher at Samsung
Publications - 41
Citations - 578
Zhengping Ji is an academic researcher from Samsung. The author has contributed to research in topics: Artificial neural network & Sparse approximation. The author has an hindex of 14, co-authored 41 publications receiving 560 citations. Previous affiliations of Zhengping Ji include Toyota & Carnegie Mellon University.
Papers
More filters
Patent
Neuromorphic spatiotemporal where-what machines
TL;DR: In this paper, a unified compact spatio-temporal method that provides a process for machines to deal with space and time and deal with sensors and effectors is described, as well as a set of additional apparatus, systems, and methods.
Proceedings ArticleDOI
Where-what network 1: “Where” and “what” assist each other through top-down connections
TL;DR: The results of the experiments showed how one type of information assists the network to suppress irrelevant information from background or irrelevant object information so as to give the required missing information in the motor output.
Proceedings Article
Radar-vision fusion for object classification
Zhengping Ji,Danil V. Prokhorov +1 more
TL;DR: Though the proposed object classification system is more flexible in terms of variety of classification tasks, the system currently demonstrates its high accuracy in comparison with others on real-world data of a two-class recognition problem.
Patent
Method and apparatus of neural network based image signal processor
TL;DR: In this article, an image signal processing (ISP) system is described, which includes a neural network trained by inputting a set of raw data images and a correlating set of desired quality output images; the neural network including an input for receiving input image data and providing processed output.
Patent
Label-free non-reference image quality assessment via deep neural network
TL;DR: In this paper, a method for training a neural network to perform assessments of image quality is provided, which includes inputting into the neural network at least one set of images, each set including an image and at least a degraded version of the image, performing comparative ranking of each image in the at least 1 set of image images, and training the network with the ranking information.