• Classified by Topic • Classified by Publication Type • Sorted by Date • Sorted by First Author Last Name • Classified by Funding Source •
Towards a Real-Time, Low-Resource, End-to-end Object Detection Pipeline for Robot Soccer.
Sai Kiran Narayanaswami,
Mauricio Tec, Ishan Durugkar, Siddharth Desai, Bharath Masetty, Sanmit
Narvekar, and Peter Stone.
In RoboCup 2022:: Robot World Cup XXV,
March 2023.
This work presents a study for building a Deep Vision pipeline suitable for theRobocup Standard Platform League, a humanoid robot soccer tournament.Specifically, we focus on end-to-end trainable object detection for effectiveperception using Aldebaran NAO v6 robots. The implementation of such a detectorposes two major challenges, those of speed, and resource-effectiveness withrespect to memory and computational power. We benchmark architectures using theYOLO and SSD detection paradigms, and identify variants that are able to achievegood detection performance for ball detection, while being able to perform rapidinference. To add to the training data for these networks, we also create adataset from logs collected by the UT Austin Villa team during previouscompetitions, and set up an annotation pipeline for training. We utilize theabove results and training pipeline to realize a practical, multi-class objectdetector that enables the robot’s vision system to run 35 Hz while maintaining
@InProceedings{RoboCup2022-nskiran,
author = {Sai Kiran Narayanaswami and Mauricio Tec and Ishan Durugkar and Siddharth Desai and Bharath Masetty and Sanmit Narvekar and Peter Stone},
title = {Towards a Real-Time, Low-Resource, End-to-end Object Detection Pipeline for Robot Soccer},
booktitle = {RoboCup 2022:: Robot World Cup XXV},
year = {2023},
month = {March},
location = {Bangkok, Thailand},
abstract = {
This work presents a study for building a Deep Vision pipeline suitable for the
Robocup Standard Platform League, a humanoid robot soccer tournament.
Specifically, we focus on end-to-end trainable object detection for effective
perception using Aldebaran NAO v6 robots. The implementation of such a detector
poses two major challenges, those of speed, and resource-effectiveness with
respect to memory and computational power. We benchmark architectures using the
YOLO and SSD detection paradigms, and identify variants that are able to achieve
good detection performance for ball detection, while being able to perform rapid
inference. To add to the training data for these networks, we also create a
dataset from logs collected by the UT Austin Villa team during previous
competitions, and set up an annotation pipeline for training. We utilize the
above results and training pipeline to realize a practical, multi-class object
detector that enables the robot’s vision system to run 35 Hz while maintaining
},
}
Generated by bib2html.pl (written by Patrick Riley ) on Wed Apr 17, 2024 18:42:57