Event Details
An Accurate and Fast Animal Species Detection System for Embedded Devices
Presenter: Mai Ibraheam
Supervisor:
Date: Fri, November 24, 2023
Time: 12:00:00 - 00:00:00
Place: ZOOM - Please see below.
ABSTRACT
Zoom Meeting Link: https://uvic.zoom.us/j/86418774095?pwd=S0s0OEVhalFRU1d1R2FOcm8xZ2RWUT09
Meeting ID: 864 1877 4095
Password: 015726
One tap mobile
+17789072071,,86418774095# Canada
+16475580588,,86418774095# Canada
Dial by your location
+1 778 907 2071 Canada
+1 647 558 0588 Canada
Note: Please log in to Zoom via SSO and your UVic Netlink ID
Abstract: For the safety of wildlife and humans, it is crucial that wildlife vehicle collisions (WVCs) and wildlife human conflicts (WHCs) mitigation systems detect wildlife in real-time. Therefore, the delay between when the wildlife first appears in the image and its detection should be enhanced, particularly on low power machines with limited computational capabilities, such as embedded systems. While most researchers propose lightweight object detectors or hardware acceleration techniques to reduce the detection delay and maximize throughput, our work takes a different approach. We aim to enhance the detection delay and power consumption by integrating two ideas into the animal species detection model while maintaining the detection accuracy: (i) the Motion-Selective Control Frames (MCF) algorithm to control the number of frames to be processed based on the motion activity within them, and (ii) two-stage pipelining-dataflow hybrid parallel processing approach. These modifications significantly reduced the processing delay and power consumption of the proposed MCFP-YOLO detector, particularly on embedded systems with limited resources, without trading off the accuracy of our animal species detection system. For field applications, the proposed MCFP-YOLO model was deployed and tested on two embedded devices, the RP4B and the Jetson Nano devices. While the Jetson Nano provided faster processing, the RP4B was selected due to its lower power consumption and a balanced cost-performance ratio, making it particularly suitable for extended use in remote areas.