------------------------------------------------------------------------------------------- 1. Our entry is based on the following PhD dissertation: - Forensic Identification by Craniofacial Superimposition using Soft Computing It was deposited and accepted for presentation by the University of Santiago de Compostela on May, 20, 2010. ------------------------------------------------------------------------------------------- 2. Names of the author, advisors, and main collaborator: -Name: Mr. Oscar Ibáñez Role: PhD author Physical address: European Centre for Soft Computing Edificio Científico-Tecnológico Calle Gonzalo Gutiérrez Quirós S/N 33600 Mieres, Asturias (Spain) E-mail: oscar.ibanez@softcomputing.es Telephone: +34 985456545, ext. 52 -Name: Dr. Oscar Cordón Role: PhD co-advisor Physical address: European Centre for Soft Computing Edificio Científico-Tecnológico Calle Gonzalo Gutiérrez Quirós S/N 33600 Mieres, Asturias (Spain) E-mail: oscar.cordon@softcomputing.es Telephone: +34 985456545, ext. 33 -Name: Dr. Sergio Damas Role: PhD co-advisor Physical address: European Centre for Soft Computing Edificio Científico-Tecnológico Calle Gonzalo Gutiérrez Quirós S/N 33600 Mieres, Asturias (Spain) E-mail: sergio.damas@softcomputing.es Telephone: +34 985456545, ext. 40 -Name: Dr. Jose Santamaría Role: main collaborator Physical address: Dpt. Computer Science, University of Jaén Edificio Tecnológico y de Ingenierías A-3 Paraje Las Lagunillas, s/n 23071 Jaén (Spain) E-mail: jslopez@ujaen.es Telephone: +34 953648585 ------------------------------------------------------------------------------------------- 3. Corresponding author: Sergio Damas (sergio.damas@softcomputing.es) ------------------------------------------------------------------------------------------- 4. Thesis abstract: Craniofacial superimposition is a forensic process that aims to identify a missing person from a photograph and a skull found. It is considered the last identification chance when other forensic techniques (fingerprints, dental records, DNA…) are not applicable. One of the crucial tasks throughout all this process is the skull-face overlay which tries to find a good fit between a 3D model of the skull and a 2D photo of the face. This craniofacial superimposition stage is usually carried out manually by forensic anthropologists. It is thus very time consuming and presents several difficulties. In this dissertation we aim to use real-coded evolutionary algorithms as a suitable approach to tackle skull-face overlay. To do so, we first formulate this complex task in forensic identification as a numerical optimization problem. Then, we adapt four different evolutionary algorithms to solve it: two variants of a real-coded genetic algorithm, the scatter search memetic algorithm, and the state-of-the-art evolution strategy CMA-ES. Besides, fuzzy sets theory has been used to deal with the different sources of uncertainty present in the problem. Results on nine superimposition problems of real-world identification cases solved by the Physical Anthropology lab at the University of Granada (Spain) are considered to test our proposals. ------------------------------------------------------------------------------------------- 5. List of satisfied criteria: A, B, E, F, G. ------------------------------------------------------------------------------------------- 6. Statement stating why the result satisfies the previous criteria: Craniofacial superimposition is a forensic process that aims to identify a missing person from a photograph and a skull found. It is considered the last identification chance when other techniques (fingerprints, dental records, DNA…) are not applicable. By projecting photographs of the skull and of the disappeared person on top of each other (or, even better, matching a scanned three-dimensional skull model against the face photo/series of video shots), the forensic anthropologist can try to establish whether that is the same person as regards the matching of some characteristic points (anthropological landmarks). Nevertheless, there is not a systematic method for the analysis by image superimposition, but every researcher applies his/her own methodology. Hence, a systematic and automatic method for craniofacial superimposition is a real need in forensic anthropology. Skull-face overlays were conducted using tracings of skulls and faces made from photographs in the thirties. In more recent times, methods have moved solely to the comparison of photographs and then to video images in the early nineties. Video-superimposition is probably now the most frequently employed method nowadays. It involves mixing two video signals of a skull and a photograph of the person in question to determine the degree of anatomical correspondence. In the last few years, the use of computers to assist the anthropologists in the identification process involved the next generation of craniofacial superimposition systems. Attempts to achieve high identification accuracy through the utilization of advanced computer technology have been a monumental task for experts in the field during the last two decades as stated by Lan [Lan, 1992]. In recent papers, some authors assert the most advanced method is based on computer-aided craniofacial superimposition through the use of the imaging tools provided by Adobe Photoshop and Corel Draw software packages [Bilge, 2003],[Al-Amad, 2006]. However, regardless the technical means used, none of the latter methods are automatic and they all assume a manual trial and error procedure, which is very time-consuming (up to 24 hours per case). In contrast, our evolutionary-based skull-face overlay method is automatic, robust, and fast (less than 4 minutes). Currently there is not an universal objective performance comparison methodology (as stated by Carl N. Stephan, "presently, it is not possible to draw firm statements concerning the overarching performance of superimposition methods because formal published studies on the accuracy and reliability of the methods have been infrequent, have used small samples, and have often not been replicated" [Stephan, 2009]). However, we can assert that our evolutionary-based automatic method achieves significantly accurate overlays as well as it is faster than the rest of skull-face overlay techniques, in several orders of magnitude (criteria E and F). There was just one previous contribution that performed an automatic skull-face overlay [Nickerson, 1993]. Nevertheless, it did not gain so much attention. Indeed, the performance of that computer-based method is far away from ours, as it was demonstrated in Chapter 3 of the referred PhD thesis (criterion B). In order to analyze the human-competitiveness of our proposal, we collaborated with the world-wide recognized Physical Anthropology lab at the University of Granada (Spain) headed by Dr. Miguel Botella. He has participated in more than 50 research projects and contracts. Among them, it is important to highlight that related with the identification of the remains of Christopher Columbus and his brother. Such project was founded by Atlantic Productions and Discovery Channel and Dr. Botella was the coordinator of the anthropological research. Besides, he has assisted the Identification Unit of the Chile Law Ministry and the Mexican authorities responsible for the identification of different murders in Ciudad Juárez, among many other achievements. We compared the overlays obtained by our automatic method proposed in the dissertation to the manual ones developed the forensic experts (see Chapter 6). That direct human-computer comparison was based on different real-world identification cases previously solved by the Physical Anthropology lab (Granada, Spain) in collaboration with the Spanish Scientific Police. The comparison relied on two different evaluation procedures: a visual assessment (made by the forensic anthropologists) and a numerical assessment based on a measure of the percentage of the head boundary that is not covered by the area of the projected skull, the Area Deviation Error (introduced in the dissertation). After a visual assessment the forensic anthropologists concluded that the overlays achieved by our approach were competitive with the best ones they could manually perform. In some cases, our overlays were even better than theirs. Besides, if we consider the area deviation error, we can see how our automatic method properly manages to get a good overall alignment of the skull and the face objects. In fact, in most of the cases these errors are approximately the half of the corresponding manual ones. In addition, they emphasized the short time needed to obtain the overlays in an automatic fashion (in the worst case, less than 4 minutes) as well as the accuracy of the resulting skull-face overlays (criteria E and F). In fact, the same group of forensic anthropologists recently used our method to solve a real-world identification case of a Portuguese man whose remains were found in the surroundings of the Alhambra monument in Granada, Spain, for the Spanish Scientific Police. It is clear that the proposed evolutionary-based method solves a problem of indisputable difficulty in its field, the skull-face overlay in craniofacial superimposition (criterion G). As Fenton stated, "the dynamic orientation process is a very challenging and time-consuming part of the skull-photo superimposition technique. Correctly adjusting the size and orienting the images can take several hours to complete" [Fenton, 2008]. As said, our method takes always less than 4 minutes to develop the latter task. Finally, as an important outcome of the research work made in our PhD dissertation, we have submitted a patent proposing a novel framework for computer-based craniofacial superimposition [Cordon, 2009] which focuses on the use of evolutionary algorithms to automate this complex problem (criterion A). [Al-Amad, 2006] Al-Amad, S., M. McCullough, J. Graham, J. Clement, and A. Hill (2006). Craniofacial identification by computer-mediated superimposition. J. Forensic Odontostomal 24, 47–52. [Bilge, 2003] Bilge, Y., P. Kedici, Y. Alakoc, K. Ulkuer, and Y. Ilkyaz (2003). The identification of a dismembered human body: a multidisciplinary approach. Forensic Science International 137, 141–146. [Lan, 1992] Lan, Y. (1992). Development and current status of skull image superimposition methodology and instrumentation. Forensic Science Review 4(2), 126–136. [Stephan, 2009] Craniofacial identification: method background and overview. http://www.craniofacial.com. [Nickerson, 1993] Nickerson, B., Fitzhorn, P., Koch, S., and Charney, M. (1991). A methodology for near-optimal computational superimposition of twodimensional digital facial photographs and three-dimensional cranial surface meshes. Journal of Forensic Sciences 36(2), 480–500. [Fenton, 2008] Fenton, T., Heard, A, and Sauer, N. (2008). Skull-photo superimposition and border deaths: identification through exclusion and the failure to exclude. Journal of Forensic Sciences 53(1), 34–40. [Cordon, 2009] Inventors (in signature order): Cordón, O., Damas, S., Ibáñez, O., Santamaría, J., Alemán, I., Botella, M. Patent title: Method and System for Forensic Identification by Craniofacial Superimposition. Application number: P200901732/3. Application date: 30/07/2009. Priority Country: Spain. Owning Institutions: Foundation for the Advancement of Soft Computing and University of Granada ------------------------------------------------------------------------------------------- 7. Full citation of the thesis: Title: Forensic Identification by Craniofacial Superimposition using Soft Computing Author: Oscar Ibáñez Advisors: Oscar Cordón and Sergio Damas Deposited and accepted for presentation by the University of Santiago de Compostela on May, 20, 2010 ------------------------------------------------------------------------------------------- 8. Any prize money, if any, will be divided equally among the co-authors. ------------------------------------------------------------------------------------------- 9. A statement stating why the judges should consider the entry as "best" in comparison to other entries that may also be "human-competitive." In this dissertation we have proposed different automatic methods based on evolutionary algorithms and fuzzy sets to solve the complex and uncertain skull-face overlay problem in craniofacial superimposition. The promising results achieved, confirmed by the forensic anthropologists of the Physical Anthropology lab at the University of Granada (see Chapter 6 of the dissertation), have demonstrated the suitability of our proposal. They emphasized the high accuracy of the resulting skull-face overlays as well as the short time needed to obtain the overlays in an automatic fashion. A graphical animation showing an example of our method solving one of the real-world cases tackled can be viewed in the project website (http://www.softcomputing.es/socovifi/en/home.php), link SOCOVIFI Project (http://www.softcomputing.es/socovifi/en/project.php). This work has attracted the attention of both the Spanish Science and Education Ministry and the Andalusian Innovation, Science and Industry Department, which granted us with two research projects funded with 79860€ and 122787€, respectively. Moreover, the quality of both the methodological framework and the evolutionary proposals have been recognized since they have been published in prestigious journals as ACM Computing Surveys (impact factor 9.92, ranked in the first position of 84 in the computer science, theory & methods category) and Information Sciences (impact factor 3.095, ranked in the eight position of 84 in the computer science, information systems category), respectively. Other associated publications in edited books, and national and international conferences are listed in the dissertation document, as the 3D Image Modelling (3DIM) Workshop, IEEE International Conference on Computer Vision, one of the most important and highly ranked conferences in the computer vision field. Since the time when we published our first results, some relevant forensic anthropologists have contacted us either to ask for information about the proposed technique or for the chance to use our technique in their identifications. Among them, we should recall Dr. M. Yoshino, Director of First Forensic Science Division of the National Research Institute of Police Science (Japan), Dr. V. Sivapiriya, Scientist of the Anthropology Division of the Forensic Sciences Department (India), and Mr. M.A. Capapé, vice-president of the association devoted to the identification of corpses in mass graves from the Spanish Civil War. We would also like to remark what Dr. Yoshino and Dr. Botella (world-wide recognized researchers in the area of forensic identification) said about our work: Dr. Yoshino: "As far as I know, your research group has the most advanced technique in craniofacial superimposition. I would like to express my deep appreciation for your achievements". E-mail sent to O. Ibáñez on April, 5, 2009. Dr. Botella: "They have been able to develop a new software tool to automate the important craniofacial superimposition technique used in forensic identification. That software is now a crucial step forward for this technique because it provides reliable craniofacial superimpositions with the required accuracy in a really short period of time in comparison with the results a human expert is able to achieve. Unlike a human expert who takes around 24 hours to determine the best craniofacial superimposition, the computer achieved an accurate solution in 2-3 minutes. We have used the software in some recent identification reports demanded by the Spanish scientific police. Just as an example, I can mention the identification of a Portuguese tourist who disappeared some years ago in the Alhambra (Spain) and whose remains appeared recently in the nearby mountains". Reference letter for an award nomination for Dr. O. Cordón, on April, 26, 2010.