IGS2011 Summary by Carolyne Bird, Document Examination, Forensic Science, Adelaide

Hans-Leo Teulings

IGS Website Manager (2007-2017)
Held in the warm and sunny climes of the Mexican Caribbean, the 15th IGS conference in Cancun attracted a contingent of over 50 presenters from fifteen nations across the globe: Australia, Austria, Canada, France, Germany, Israel, Italy, Japan, Mexico, The Netherlands, Spain, Switzerland, Tunisia, the United Kingdom and the United States of America.

The welcome reception in the evening of Sunday 12th June, held in the poolside area of the Live Aqua Cancun, commenced with watermelon daiquiris and a Marimba band. Although there was no formal welcome speech from the conference Chair, Jose ‘Pepe’ Contreras-Vidal (he didn’t want to interrupt the convivial atmosphere), we all felt a warm Mexican greeting.

The (unofficial) theme throughout the conference was brain-mind machines and brain-machine interfaces (BMI). The first full day of conference proceedings began with a plenary presentation by Mikhail Lebedev (Duke University Center for Neuroengineering), on the common theme. Lebedev reported outcomes of recent work with rhesus monkeys that had been implanted with multielectrode arrays in multiple cortical areas. Neuronal activity could be recorded and decoded for various motor activities, and direct stimulation of the brain encouraged movement in the monkeys (bipedal locomotion and reaching movements). This paves the way for neural prosthetic devices in the future, able to restore locomotive and fine motor movements to disabled people.
Reported research from other presenters used other, sometimes less invasive, means of detecting and transmitting brain signals for motor control, and covered various aspects relating to encoding and decoding of these processes (Ramirez, National Institute of Astrophysics, Mexico; Paek, University of Maryland, USA; Bensmaia, University of Chicago, USA; Santello, Arizona State University, USA; Valero-Cuevas, University of Southern California, USA; Crone, John Hopkins University, USA; Contreras-Vidal, University of Maryland, USA).

Aspects of human machine interfaces involving pen-based inputs were covered by Marcus Liwicki (German Research Center for Artificial Intelligence, Germany), Hala Bezine (National School of Engineers of Sfax, Tunisia), and Peiyu Li (Université Européenne de Bretagne, France).

Plenary speaker Marc H. Schieber (University of Rochester School of Medicine & Dentistry) gave an easy to understand (even to someone to whom this is all relatively unfamiliar) presentation on the neuroscience of finger movements and his research into whether motor cortex neurons can be combined into small groups separated from finger movements, and used to drive a cursor. Other research dealing with movement neuroscience was reported in two sessions over the three days and included asymmetry of bilateral transfer (van Gemmert, Louisiana State University, USA), characterising movement synergies of grasping in the human hand (Agashe, University of Maryland, USA), and presentations investigating movement/motor control features of Parkinson’s disease (Chong, Georgia Health Sciences University,USA; Prashad, University of Maryland, USA).

The third plenary speaker, Ranulfo Romo (National Autonomous University of Mexico, Mexico) discussed neural activity without the movement aspect, aiming in his research for an integrated understanding of how subjective sensory experience arises in the activity of the brain.

Moving completely away from the neural side of the equation, there was a session on computational models and two on handwriting analysis and technology. Presentations within these covered artificial neural networks for virtual robotic anthropomorphic arm movements (Beltrán-Blanco, Polytechnic University of Cartagena, Spain), a feedback loop for improving writing order recovery of off-line cursive handwriting (Senatore, University of Salerno, Italy), an algorithm for segmenting isolated characters within cursive words (Marcelli, University of Salerno, Italy), generation of synthetic handwritten gestures using sigma-lognormal model (O’Reilly, Ecole Polytechnique de Montreal, Canada), deterioration of handwriting processes and its association with decreasing executive function (Rosenblum, University of Haifa, Israel), and the handwriting of children with autism, Asperger’s disorder and attention deficit hyperactivity disorder (Johnson, Monash University, Australia).

As I come to the IGS with a forensic handwriting examination background, I was most interested in the forensic-related presentations, which covered three sessions over the three days. Two presentations dealt with differentiating scribes: Claudio de Stefano (University of Cassino, Italy), developed a pattern recognition system using digital image processing techniques to characterise (with at least 91% accuracy) the features & habits of individual scribes who worked together in the transcription of a medieval single book. Martin Jarvis (Charles Darwin University, Australia) applied forensic handwriting examination methods to some of the manuscripts of both Johann Sebastian Bach and Anna Magdalena & found that their music-calligraphies can be differentiated. Further research by Jarvis investigated the contention that if Anna Magdalena wrote with a quill cut by Johann Sebastian, then it would be more likely that her music-calligraphy would appear similar to his. Findings do not support this hypothesis, but suggest that the use of the quill-pen in no way results in features that are fundamentally dissimilar to those that result from the scribes’ normal movement behaviour.

Forensic related presentations investigating computer-based methods for handwriting and signature verification, identification and classification were made by Reinier Verduijn (Netherlands Forensic Institute, the Netherlands) and Richard Guest (University of Kent, UK).

Emily Will (QDE Will, USA) devised a study to investigate FHEs skill, and the features they used, for inferring handwriting speed from a static image. Results showed that FDEs can make correct inferences of handwriting speed from the static trace. I (Bird et al., LaTrobe University, Australia) also presented an investigation into some dynamic features of handwriting, reporting findings of a study into the dynamic differences between naturally written, disguised and forged handwritten text. Results suggest that in theory, velocity may be used to discriminate between disguised and forged handwriting behaviour. Mariangela Genna (University of Trieste, Italy) also examined kinematic features of handwriting and how these change in children as schooling proceeds, concluding that some of these parameters may be useful for handwriting development studies. A study of eye movements and handwriting dynamics while subjects attempted to simulate two model signatures of different complexities was undertaken and reported by Avni Pepe (LaTrobe University, Australia).

The poster presentations, while not extensive (about a dozen in total) were diverse in their topics, from remedial handwriting programs to aspects of automated signature verification.

The Tuesday night banquet on the beach around sunset was a lovely event and gave attendees the chance to mingle in a social atmosphere, which may have been a little absent at other times (many participants, myself included, brought their spouse or family along to sunny Mexico for a holiday and thus spent time outside of the sessions with them, rather than networking). The Mariachi band was lively and entertaining, and the promise of turtles laying eggs on the beach enticed a number of people to stay up late to catch a glimpse.

At the conclusion of the conference, the best student paper awards were presented by a representative from the sponsoring organisation to the following students:

Forensic Sciences by the Association of Forensic Document Examiners

Avni PEPE, A cognitive look into simulations of high and low complexity signatures
Cognitive-Motor Neuroscience by the University of Maryland

Christian O'REILLY, The coupling of agonist and antagonist commands in speed/accuracy tradeoff
Neural Decoding or BMI by g.tec Medical Engineering

Shikha PRASHAD, The macrostructure of rest tremor in Parkinson's disease
Computational Model or Neuroscience by the University of Cartagena

Andrew PAEK, Decoding finger movements from brain activity acquired via scalp electroencephalography
Graphonomics by IGS

Rosa SENATORE, From motor to trajectory plan: a feedback loop between unfolding and segmentation to improve writing order recovery

In all, I found the scientific presentations interesting and informative. I valued the opportunity to present some of my own research to and receive feedback from such esteemed company.

By: Carolyne Bird, Document Examination Section, Forensic Science South Australia, Adelaide, Australia