US Patent Issued to Intel on June 2 for "Per kernel KMeans compression for neural networks" (Israeli, American Inventors)

ALEXANDRIA, Va., June 2 -- United States Patent no. 12,645,923, issued on June 2, was assigned to Intel Corp. (Santa Clara, Calif.).

"Per kernel KMeans compression for neural networks" was invented by Yonatan Glesner (Kfar Haro'e, Israel), Gal Novik (Haifa, Israel), Dmitri Vainbrand (Sunnyvale, Calif.) and Gal Leibovich (Kiryat Yam, Israel).

According to the abstract* released by the U.S. Patent & Trademark Office: "Methods and apparatus relating to techniques for incremental network quantization. In an example, an apparatus comprises logic, at least partially comprising hardware logic to determine a plurality of weights for a layer of a convolutional neural network (CNN) comprising a plurality of kernels; organize the plurality of weight...