Smith, V. A. (2015). The use of 3D sensor for computer authentication by way of facial recognition for the eyeglasses wearing persons (Doctoral dissertation, Colorado Technical University).
For this entry, I am looking at a Vincente A. Smith’s dissertation-length research report in the field of systems engineering, which attempts to explore the restrictions of cost-effective facial recognition for glasses-wearing users.
Smith’s report is highly technical, but essentially studies computer-based facial authentication through sample-based study of real-world 3d-systems-based PC access. Using these results, it then attempts to address a primary research question and multiple sub-questions:
1) How 3d sensors might aid in glasses-inhibited facial identification for personal users; and 2) what the merits are of 3d sensing over traditional technologies for users, 3) how detection rates have improved over time, 4) what requirements reduce false positives/negatives, and 5) what improvements are needed to hardware in future iterations (76-8)?
Through literature review, research, and metaanalysis, Smith attempts through these questions to establish improved low-cost methods for resolving authentication and access concerns for eyesight-impaired users and to provide recommendations for future development. To the extent that his research argues for a resolution of this issue, this argument is supported by existing research which demonstrates the limitations of these technologies in current professional, personal, and governmental contexts and the need for new processes which increase the viability of access for all users.
Smith’s literature review is comprehensive and serves to support his research goals and delimit his research processes by building upon previous reporting; additionally, his methodology is sound and well-documented, if limited by various significant factors. It is these factors – and Smith’s resolution of the challenges they present, which make this research useful to my general studies.
The research grapples with multiple limiting factors, each of which are exacerbated by the researcher’s graduate student status and limited research funding and access: hardware (83, 101) and software (47) availability, cost, and support (as well as laboratory security); sample size, sourcing, and privacy (80, 99); expertise restrictions (101); and time restrictions (36).
The narrative of Smith’s resolution of these issues has proven an invaluable resource when extrapolating qualitative research narratives regarding graduate student austerity limitations from research reporting. I selected this as a case study even though Smith is outside the humanities: a choice made because Smith’s reporting is frequently (and remarkably) transparent regarding the restrictions of austerity upon his methodological/reporting choices.
Given this transparency in a broad (and technologically complicated) research process, Smith has proven an excellent resource for studying the limitations (and workarounds) of austerity-contextualized work. Smith’s deployment of F/OSS solutions reflects several arguments from Zoetewey’s “The Rhetoric of Free” regarding how researchers navigate financial limitations through deployments of counter-strategies which necessitate inferior/non-standardized results. By contrast, this dissertation also serves to demonstrate categorization blind spots in Aksulu and Wade’s research synthesis; it, and much research like it, runs parallel with and intersects F/OSS research concerns – but does not name or address such paradigms and would challenge categorization in their model.
Due to its technical specificity, while this research is specifically useful within trends I am studying I would not recommend it for general F/OSS studies in the humanities.
(Body length – 500 words)