MARCS Monday Meeting (MMM): Werrington Edition

Event Name MARCS Monday Meeting (MMM): Werrington Edition
Start Date 30th Sep 2019 11:00 am
End Date 30th Sep 2019 12:00 pm
Duration 1 hour
Description
Please join us at the next MMM – Werrington South edition for Saeed Afshar and Andrew Nicholson's presentation on event based cameras.

Venue: Level 2, Building BA, (Large meeting space), Werrington South Campus

Speaker: Dr Andrew Nicholson

Topic: Event-based Cameras

Abstract: Event-based cameras are a light sensor array inspired by evolution – a design that attempts to mimic the biological architecture of the retina. Within each pixel, relative changes in light intensity over time trigger events independently of other pixels, as contrasted with conventional frame-based cameras. Events are either an increase or decrease in temporal contrast for an individual pixel sensor within the array.

Compared to traditional frame-based cameras, event-based cameras have large dynamic range, high temporal resolution, low latency and low power consumption, and as such are considered suited to embedded sensing, robotics and wearable devices.

ICNS is fabricating our own designs of these pixel sensors over the next years to study their properties and uses in various scenarios including event tracking and object detection in space, and interfacing these event streams to a neuromorphic signal processor and/or spiking neural networks.

Speaker: Saeed Afshar

Topic: High Speed Event-based Visual Processing in the Presence of Noise

Abstract: Standard machine vision approaches are challenged in applications where large amounts of noisy temporal data must be processed in real-time. This work aims to develop neuromorphic event-based processing systems for such challenging, high-noise environments.

The novel event-based application-focused algorithms developed are primarily designed for implementation in digital neuromorphic hardware with a focus on noise robustness, ease of implementation, operationally useful ancillary signals and processing speed in embedded systems. By probing the large search space of potential hardware solutions to real-time temporal data processing in high-noise applications, this work seeks to develop functionally equivalent cognitive processes as they occur in the brain in the same decentralized noise-robust manner and, importantly, through the use of time itself as the central processing variable.


IMPORTANT
Please note, there will be no MMM or MAC on 7th October, 2019 due to the public holiday.

We thank you for your continued attendance at the MMM's.  Let's keep the momentum going.

MARCS staff and students are reminded that all meetings and workshops have an important role in building and maintaining the sense of community which is central to the success of MARCS as a cooperative and energetic research institute.  Your attendance is both welcomed and expected.

The zoom ID is: 627 146 998. Link to zoom https://uws.zoom.us/j/627146998