{"schema_version":"1.0","canonical_url":"https://patentable.app/patents/US-11521074","patent":{"patent_number":"US-11521074","title":"Flexible, lightweight quantized deep neural networks","assignee":null,"inventors":[],"filing_date":"2020-05-29T00:00:00.000Z","publication_date":"2022-12-06T00:00:00.000Z","cpc_codes":["G06N","G06N","G06N","G06N","G06N"],"num_claims":18,"abstract":"To improve the throughput and energy efficiency of Deep Neural Networks (DNNs) on customized hardware, lightweight neural networks constrain the weights of DNNs to be a limited combination of powers of 2. In such networks, the multiply-accumulate operation can be replaced with a single shift operation, or two shifts and an add operation. To provide even more design flexibility, the k for each convolutional filter can be optimally chosen instead of being fixed for every filter. The present invention formulates the selection of k to be differentiable and describes model training for determining k-based weights on a per-filter basis. The present invention can achieve higher speeds as compared to lightweight NNs with only minimal accuracy degradation, while also achieving higher computational energy efficiency for ASIC implementation."},"analysis":{"summary":null,"layman_explanation":null,"technical_analysis":null,"business_analysis":null,"faqs":null,"topics":[],"tech_cluster":null},"seo":{"title":"Flexible, lightweight quantized deep neural networks","description":"To improve the throughput and energy efficiency of Deep Neural Networks (DNNs) on customized hardware, lightweight neural networks constrain the weights of DNNs to be a limited combination of powers o","keywords":[]},"attribution":{"source":"Patentable","source_url":"https://patentable.app","canonical_url":"https://patentable.app/patents/US-11521074","license":"CC-BY-4.0-like","license_terms":"AI-generated analysis on this page (summary, layman_explanation, technical_analysis, business_analysis, faqs) may be reused with attribution and a visible link back to the canonical URL above. Patent abstracts, claims, and bibliographic data are USPTO public domain.","required_link":"https://patentable.app/patents/US-11521074","citation_suggestion":"Patentable. \"Flexible, lightweight quantized deep neural networks\" (US-11521074). https://patentable.app/patents/US-11521074","copyright_holder":"Nomic Interactive Technology LLC"},"links":{"html":"https://patentable.app/patents/US-11521074","json":"https://patentable.app/api/llm-context/US-11521074","site":"https://patentable.app","llms_txt":"https://patentable.app/llms.txt"},"generated_at":"2026-05-31T10:40:28.271Z"}