In this dataset, vector quantization claims mostly originate in video and audio codec work, covering techniques like product quantization, residual quantization and codebook training that compress signal or vector data while preserving the ability to approximate distances between compressed representations. That same machinery is increasingly repurposed for retrieval systems, where compressed embedding vectors need to be compared quickly at query time. The IPC composition shows H04N and G10L still account for the largest shares of records, meaning most granted claim territory frames the problem as media compression rather than search infrastructure.

The ranking covers 100 companies, with a single leader holding 238 records — more than four times the count at fifth place, which sits at 59. The top 5 assignees together hold 44.9% of all 1,327 records in scope, and the top 10 extend that to 57.6%, so the field is genuinely concentrated at the top rather than evenly spread. Company names in the underlying data span consumer electronics, telecom equipment and codec-standards research organisations.

Filings peaked at 73 in 2018 and have since eased, with 2021 at 43 and 2024 at 32 — a -26% change over that three-year span, which is the most recent period reliable enough to compare. Figures for 2025 and 2026 look lower still, but publication typically lags actual filing by around 18 months, so those years are undercounted rather than necessarily reflecting reduced activity. A clearer read on the current trajectory will only be available once those years finish publishing.