Subprojects

The project has few subprojects. Some of them are developed in parallel and others are part of the end of the global project

• Smart Imagers
• Acoustic Sensors
• 3D Arquitectures
• Algorithms
• Field Tests

Last update on February 10th, 2013.

The Project

The following is a brief introduction of each of the subprojects that comprise the whole project. Each one has its own page where datasets, papers and progress of the project are listed.

Smart Imagers

Within this line of research we propose to continue with the objective of building a complete smart vision system within the same chip. This requires the development of three different hierarchies of processing. The lowest level hierarchy is composed of the pixel array itself. The intermediate level of hierarchy is composed of operations over the array, that can be used to extract, with some more processing, features of the images. The highest level of hierarchy is composed of a built-in microprocessor able to coordinate the array, the intermediate processing units, and the available memory.

Acoustic Sensors

The acoustic processor will provide another level of redundancy by locating targets based on bearing and range estimations. Systems based on interaural time difference (ITD) and micro-doppler will be investigated, including modifications for the identification of multiple targets.

3D Arquitectures

We aim at gaining experience on 3D architectures by studying three different systems on which we have an extensive experience, like particle filters, visual and acoustic neural processors and piecewise linear computation structures. It is necessary to develop test structures and a design flow.

Algorithms

In autonomous vehicle applications each resource, a node, need to exchange observations and data to estimate their pose. Also it is important that each node be able to predicts, not only its own pose, also other nodes ones. Due to bandwidth restrictions, node capacity (hardware and power) and perception needs, this is not a simple task.

Field Tests

The experimental scenario is a fundamental part of the project as it will permit to properly test its hypothesis: a reliable resource localization tool for autonomous applications. We have several places that can be used as a test bed. Of particular interest it is a facility of the University of Sydney, Australia where vehicles can be retrofitted with a communication platform for safety purposes.