Search

ABSTRACTFace identification systems that operate on long wave infrared (LWIR) images are able to overcome some of the limitations of approaches based on visible images, such as dealing effectively with illumination variations. Nonetheless, distortions of perceptual image quality can impair the performance of thermal face recognition systems. Although the interaction between perceptual image quality and tasks such as face detection has been studied on visual images, the development of similar models has not been applied to the LWIR-based face recognition problem. Here, we analyze the impact of four common infrared image distortions (gaussian noise, blur, non-uniformity, and JPEG compression) on two thermal face recognition systems. We propose an LWIR image face recognition framework, based on thermal signature templates, and enhanced by natural scene statistics image quality descriptors, which achieves system robustness against image quality distortions. Furthermore, we develop a novel infrared face recognition system that is based on the complex wavelet structural similarity (CW-SSIM) index, which exhibits resistance to image distortions, within a relatively simple implementation. Our results validate the applicability of image quality assessment models to biometric tasks on LWIR images. To facilitate our study, we created two new LWIR facial image databases, with different poses, expressions and illumination conditions.

As the production, the variety, and the consumption of borndigital video grows, so does the demand for acquiring, curating and preserving large-scale digital video collections. A multidisciplinary team of curators, computer scientists and video engineers we explore the use of Non-Reference Image and Video Quality Algorithms (I/VQA), specifically of BRISQUE in this paper, to automatically derive ranges of video quality. An important characteristic of these algorithms is that they are modeled to human perception. We run the algorithms in a High Performance Computing (HPC) environment to obtain results for many videos at the same time, accelerating time to results and precision in computing per-frame and per-video quality assessment scores. Results, which were evaluated quantitatively and qualitatively, suggest that BRISQUE identifies the distortions in which it was trained, and performs well in videos that have natural scenes and do not have drastic scene changes. While we found that this particular model is not apt for evaluating collections with varied content, the results suggest that research into other I/VQA models is promising, and that their implementation at large scale can narrow the problem of curating very digital video collections and lead to preservation and access decisions based on informed priorities. [ABSTRACT FROM AUTHOR]