US Patent Issued to Baidu USA on Nov. 4 for "Automatic channel pruning via graph neural network based hypernetwork" (California Inventors)

ALEXANDRIA, Va., Nov. 6 -- United States Patent no. 12,462,157, issued on Nov. 4, was assigned to Baidu USA LLC (Sunnyvale, Calif.).

"Automatic channel pruning via graph neural network based hypernetwork" was invented by Baopu Li (Santa Clara, Calif.), Qiuling Suo (Sunnyvale, Calif.) and Yuchen Bian (Santa Clara, Calif.).

According to the abstract* released by the U.S. Patent & Trademark Office: "Model pruning is used to trim large neural networks, like convolutional neural networks (CNNs), to reduce computation overheads. Existing model pruning methods mainly rely on heuristics rules or local relationships of CNN layers. A novel hypernetwork based on graph neural network is disclosed for generating and evaluating pruned networks. A graph...