Panasonic’s Face Matching technology has ranking highly in benchmark testing by the US National Institute of Standards and Technology, for the second year running. In co-operation with IARPA (Intelligence Advanced Research Projects Activity), NIST tested a forthcoming version of Panasonic’s facial recognition technology, which is currently being developed at the company’s Innovation Centre. The benchmark test consists of eight challenges addressing verification, identification, detection, clustering, and processing of crowded images.
The benchmarking is carried out with a pool of 67,000 face images, 700 face videos, and 10,000 non-face images, involving the comparison of miss and hit rates highlighted by each respective algorithm. The Panasonic algorithm came out top of the benchmark for cumulative match characteristic – the fraction of hits above a certain threshold of accuracy
The test is often performed with photo-journalism images, due to their wide variations in pose, illumination, expression, and occlusion which creates a deliberate challenge. “The NIST result is excellent news and is yet another demonstration of Panasonic’s commitment to adding additional value and real quality within our range of security technologies,” said Gerard Figols, Panasonic’s Security Solutions Product Marketing Manager. “Panasonic has been the world’s number one patent producer for each of the last three decades, and we place a great amount of value on innovation and research and development