III Percent Patriots: Rightful liberty is unobstructed action according to our will within limits drawn around us by the equal rights of others. ~ Thomas Jefferson


Click the Image

Dec 6, 2011

Solo Act

Sometimes folks just like to be left alone to do their own thing.

The Tree of Liberty has a post here on what the soothsayers of DARPA are doing.

Well is seems now they want to use their fairy dust to see how the recluse, hermit or misfit will bomb the USG.


Straight from the Jackass's mouth
Program:  SBIR
Topic Number:  SB121-002 (DARPA)
Title:  Assessment of Asymmetric Social Indicators




Objective:  Develop and implement algorithms for capturing, measuring, and assessing social interactions in military contexts (training, medical and mental health settings, security) based on asymmetric data.
  Description:  Effective social interactions are vital in military, law enforcement, and other life critical environments. Practitioners under extreme stress are routinely asked to enter unfamiliar situations in cultures alien to their own and conduct activities, adjudicate arguments, and facilitate life altering decisions. Enabling these complex social interactions affords improved performance (e.g., fewer misunderstandings, increased efficiency, better decisions). However, in order to train or provide feedback to individuals about their social interactions, one must first have an understandable way to capture and represent those interactions. Manual, ethnographic techniques provide significant insights, but are not scalable. Automated techniques involving wearable sensors and video processing show significant promise. However, these methods assume that all parties involved in an interaction are instrumented in a way that allows for subsequent data analysis. In many realistic situations, it is only possible to instrument one party in an interaction. For instance, military emergency department staff involved in an on-the-job assessment could be instrumented while their patients could not. This limits the environments in which these technologies can be actively used to training and experimental venues, which lose real-world contextual cues critical in social interactions. This topic seeks innovative algorithmic approaches to "filling in the gaps" in social interactions when data collection systematically misses one party in bi-lateral and multi¬-lateral social encounters. Solutions should show the utility of particular approaches on quasi-realistic data sets, and should articulate the assumptions being made respecting data collection approaches, sensor types, and domain requirements. A field or laboratory validation study is critical to assess the utility of any proposed approach. Solutions must be capable of hypothesizing social interactions with 70% accuracy compared to "ground truth" in a “known” realistic dataset provided by the performer. Alternative metrics of performance will be considered. Algorithms must be capable of being run on commercially available PCs and, eventually, state of the art mobile platforms. Interpretability by relevant subject matter experts of the hypothesized social interactions must be demonstrated.
  PHASE I: Describe and implement algorithms designed to represent bi-lateral or multi-lateral social interactions based on datasets chosen by the company with incomplete or asymmetric collection. Demonstrate utility of the algorithms in a quasi-realistic dataset that does not involve human use to generate.
  PHASE II: Refine and implement algorithms; integrate algorithms within a system created by the performer designed for description, measurement, and assessment of social interaction or team performance; demonstrate on a realistic dataset.
  PHASE III: Develop operational system for use in assessing military and law enforcement interactions in operational settings. Should be able to examine effectiveness of military or law enforcement training oriented toward social performance with asymmetric data collection.

No comments:

Post a Comment