Thoughts on object detection using convolutional neural networks for forward-looking sonar

Livne, Assaf; Baruch, Alon; Guterman, Hugo

doi:10.15406/iratj.2018.04.00120

Cited by 4 publications

(1 citation statement)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The FLS is frequently used by AUVs as an object detection, and obstacle avoidance sensor [17], [18], [19], [20]. In [21], The term bottleneck refers to the fact that the network typically reduces the dimensionality from that of the image space to the feature space. In the context of transfer learning the hope is that similar features can be used for different learning tasks.…”

Section: Introductionmentioning

confidence: 99%

Object Recognition in Forward Looking Sonar Images using Transfer Learning

Fuchs

Gällström²,

Folkesson

2018

2018 IEEE/OES Autonomous Underwater Vehicle Workshop (AUV)

View full text Add to dashboard Cite

Forward Looking Sonars (FLS) are a typical choice of sonar for autonomous underwater vehicles. They are most often the main sensor for obstacle avoidance and can be used for monitoring, homing, following and docking as well. Those tasks require discrimination between noise and various classes of objects in the sonar images. Robust recognition of sonar data still remains a problem, but if solved it would enable more autonomy for underwater vehicles providing more reliable information about the surroundings to aid decision making. Recent advances in image recognition using Deep Learning methods have been rapid. While image recognition with Deep Learning is known to require large amounts of labeled data, there are data-efficient learning methods using generic features learned by a network pre-trained on data from a different domain. This enables us to work with much smaller domain-specific datasets, making the method interesting to explore for sonar object recognition with limited amounts of training data. We have developed a Convolutional Neural Network (CNN) based classifier for FLSimages and compared its performance to classification using classical methods and hand-crafted features.

show abstract

Section: Introductionmentioning

confidence: 99%

Object Recognition in Forward Looking Sonar Images using Transfer Learning

Fuchs

Gällström²,

Folkesson

2018

2018 IEEE/OES Autonomous Underwater Vehicle Workshop (AUV)

View full text Add to dashboard Cite

show abstract

Autonomous underwater vehicles: Instrumentation and measurements

Sánchez¹,

Papaelias

Márquez³

2020

IEEE Instrum. Meas. Mag.

View full text Add to dashboard Cite

Oceans exploration and inspection are being a great challenge for the industry nowadays. The underwater instrumentations and measurements are being improving due to the current technologies, or by development new ones, to cover the demand of the new industry offshore. The Autonomous Underwater Vehicles, a subcategory of submarine, is used to perform the subaquatic tasks. This vehicle provides advantages for underwater works, e.g. safety and reliability inspections, but it also offers limitations as sensors systems, monitoring and communications systems, autonomous operational endurances, propulsion systems or mapping designs, etc. The main scientific contributions of this paper are: An review of the state of art in novel and main instrumentation and measurement systems embedded in AUVs; It is showed the future uses and development; The paper synthesis the main and current navigation, mapping and sampling technologies, together with different applications.

show abstract