Background:

During the 8:30am, 10:00am, and 11:00am, sessions pay close attention to the 5-10 minute project reports. Whichever ones most interest you should correspond to the afternoon sessions at 1:00pm and 3:15pm. Have your agenda with you so that you can take notes on which afternoon sessions you want to attend. Decide during lunch and the breaks in between sessions how to split yourselves up into the groups so that no more than 3 of you are in any 1 group in the 1:00pm session, and no more than 4 of you are in any 1 group in the 3:15pm session. You can choose who attends which session however you want.

In the afternoon sessions you choose to visit, play close attention and take good notes! This blog assignment should be treated like a “report back” to your colleagues who aren’t able to attend the others. You have to report the information as accurately and completely as you are able to. Some questions to consider for each of the afternoon sessions:

  1. What forensic methods are discussed?

For session 1 the forensic methods that were discussed were ways to create procedures on how to collect evidence at a crime scene and for session 2 the only method that was discussed was the job of a forensic document examiner.

  1. What statistical methods are discussed?

For session 1 the statistical methods that were discussed were likehood ratios, bayes theorem and algorithms. For session 2 the method that was discussed was a software called MovAlyzer which helps collect the time, position and pressure of someone’s handwriting.

  1. If you were a full-time researcher with CSAFE, which one of the projects presented in the group would you most like to work on? Why?

If I were a full-time researcher with CSAFE one project I would love to work on is Research on Lawyers, Jurors, and the Evaluation of Forensic Evidence because in court the judge and jury are the lifeline between a person going to jail or not. There have been cases where a innocent person was sent to jail due to the judge, lawayer, and jury not being more open-minded when it comes to the forensic evidence, they base decisions on word of mouth and not on the evidence presented to them, this is where the statistical models comes in and we provide the numbers to accurately convict the right people.

  1. What future work is discussed? What holes are the projects trying to fill?

For session 1 the future work they want done is to create a step by step process for collecting forensic evidence and to make sure to correctly figure out what is relvant to the crime itself, for session 2 the future work they discussed was to improve statistical models so more accurate readings can be determined.

  1. What future work could you propose to the group that would also be of interest?

More collaberation amongst once another so we can make sure the forensic evidence we collect wether it is a bullet case or shoeprint is pointed to the correct person and with more statistical evdience being created there are higher chances to prove forensic evidence credible in the courtrooms.

  1. What was the coolest or most intriguing thing you learned in the session?

The coolest thing I learned in both sessions was I was so amazed of how many brilliant people can come together and share their knowledge amongst one another. Everyone in each sessions had great ideas on how to imporve the accuracyof forensic evidence and watching presentations and all the questions being asked has gave me so much knowledge and makes me love being a forensic science major .

Poster Summary

The poster I did my review on was “A Probabilistic Model for Shoeprint Accidentals” by Neil Spencer and Jared Murray but only Neil was there to present. They are both from Carnegie Mellon University and they really didn’t have a research question but more of a objective which was to develop a probabilistic model for the locatins of accidentals on the sole of a shoe. The model helps quantify the rarity of the suspect’s shoe as being a crime scene print. One result they report on is that the current model that work on these shoeprints also is comparably to another model which only uses contact surface information, basically he stated there is more room for improvement for this research and one question I asked him “ How does the Kernel increase the intensity of the shoe? Does is help show the shoeprint in more clarity?” I enjoyed speaking with Neil and I cannot wait to learn more about his research when it’s finally completed.

Bonus Challenge: Picture with Neil Spencer

me with researcher Neil Spencer

1pm Session: Human Factors

This session was a very interesting one, this session was all about how human factors plays a huge part when it comes to dealing with forensic evidence. There was many questions involving on how do we prove forensic evidence if the jury or judge just go by the word of mouth so many people issued great ideas such as creating a process for forensic processing including a step-by-step guideline or even a diagram but then as the conversation escalated it started to move away from a discussion to more of a informative arguement, but that can happen when great minds are in one room together they end up clashing. It all calmed down later on and then they came back together and sugguested the jury, judge, and lawyers get training on forensic evidence so that way they can become more comfortable with what is presented to them. Overall, I would say this was my favorite session because the court plays a huge part with forensic evidence and if people stay educated or familiarize there will be less chances of innocent people going to prison.

3:15pm Session: Firearms/Handwriting

Session 2 was all about firearms and handwriting forensic evidence, the first part of the session Hal Stern talked about how he can determine different aspects of handwriting thanks to a statistical model called MovAlyzer software in which it helps collect the time, position and pressure of something handwritten. Then more people gave their opinion and asking is there more ways to improve accuracy of something handwritten and of course there is he said but we as researchers have to go out and develop them. For the second part of the session they next talked about bullet casings and cartridge and how statistical models can help find matching striae on both the bullet and cartridges, they also talked about how one model uses 3D images to help show better clarity on the bullets. It was a very informative session and each researcher gave tips to one another on how to help better and further their research. It is very amazing to see intelligent people share their expertise with one another.

Poster Session: A Probabilistic Model for Shoeprint Accidentals

At the Mixer and Poster Session I went and talked with Neil Spencer about his research on how he developed a model which provides a data-driven method for quantifying the rarity of a accidentals on a particular shoe, basically this model allows us to quantify the uncertainty that a suspect’s shoe is the source of a crime scene print. He tries to increase the intensity of the prints by his four controlled factors which are 1. The shape and intensity of the contact surface at the location, 2. The region of the shoe of which it is a part which could be the heel or toe part of the shoe, 3. A random shoe-specific noise term, and 4. The probability of seeing accidentals at other nearby locations. He Concluded that there is of course more future work to be done due to the fact that his current model performs comparably to a model which only uses contact surface information. It was great speaking to him and learning so much about his research experience and about him as a researcher.