Recent improvements in object detection are driven by the success of
convolutional neural networks (CNN). They are able to learn rich features
outperforming hand-crafted features. So far, research in traffic light
detection mainly focused on hand-crafted features, such as color, shape or
brightness of the traffic light bulb. This paper presents a deep learning
approach for accurate traffic light detection in adapting a single shot
detection (SSD) approach. SSD performs object proposals creation and
classification using a single CNN. The original SSD struggles in detecting very
small objects, which is essential for traffic light detection. By our
adaptations it is possible to detect objects much smaller than ten pixels
without increasing the input image size. We present an extensive evaluation on
the DriveU Traffic Light Dataset (DTLD). We reach both, high accuracy and low
false positive rates. The trained model is real-time capable with ten frames
per second on a Nvidia Titan Xp.
%0 Generic
%1 citeulike:14607463
%A xxx,
%D 2018
%K detection finegrained loss ssd
%T Detecting Traffic Lights by Single Shot Detection
%U http://arxiv.org/abs/1805.02523
%X Recent improvements in object detection are driven by the success of
convolutional neural networks (CNN). They are able to learn rich features
outperforming hand-crafted features. So far, research in traffic light
detection mainly focused on hand-crafted features, such as color, shape or
brightness of the traffic light bulb. This paper presents a deep learning
approach for accurate traffic light detection in adapting a single shot
detection (SSD) approach. SSD performs object proposals creation and
classification using a single CNN. The original SSD struggles in detecting very
small objects, which is essential for traffic light detection. By our
adaptations it is possible to detect objects much smaller than ten pixels
without increasing the input image size. We present an extensive evaluation on
the DriveU Traffic Light Dataset (DTLD). We reach both, high accuracy and low
false positive rates. The trained model is real-time capable with ten frames
per second on a Nvidia Titan Xp.
@misc{citeulike:14607463,
abstract = {{Recent improvements in object detection are driven by the success of
convolutional neural networks (CNN). They are able to learn rich features
outperforming hand-crafted features. So far, research in traffic light
detection mainly focused on hand-crafted features, such as color, shape or
brightness of the traffic light bulb. This paper presents a deep learning
approach for accurate traffic light detection in adapting a single shot
detection (SSD) approach. SSD performs object proposals creation and
classification using a single CNN. The original SSD struggles in detecting very
small objects, which is essential for traffic light detection. By our
adaptations it is possible to detect objects much smaller than ten pixels
without increasing the input image size. We present an extensive evaluation on
the DriveU Traffic Light Dataset (DTLD). We reach both, high accuracy and low
false positive rates. The trained model is real-time capable with ten frames
per second on a Nvidia Titan Xp.}},
added-at = {2019-02-27T22:23:29.000+0100},
archiveprefix = {arXiv},
author = {xxx},
biburl = {https://www.bibsonomy.org/bibtex/25915c47e43c2249978160e068067fd0d/nmatsuk},
citeulike-article-id = {14607463},
citeulike-linkout-0 = {http://arxiv.org/abs/1805.02523},
citeulike-linkout-1 = {http://arxiv.org/pdf/1805.02523},
day = 7,
eprint = {1805.02523},
interhash = {c81c83b5cc06a76835e7bd556b3ee896},
intrahash = {5915c47e43c2249978160e068067fd0d},
keywords = {detection finegrained loss ssd},
month = may,
posted-at = {2018-06-25 11:34:08},
priority = {4},
timestamp = {2019-02-27T22:23:29.000+0100},
title = {{Detecting Traffic Lights by Single Shot Detection}},
url = {http://arxiv.org/abs/1805.02523},
year = 2018
}