Bi-focal unattended ground sensor concept for high resolution and wide field-of-view real-time situational awareness

Hueber, Nicolas; Pichler, Alexander; Hennequin, Christophe; Voisin, Philippe; Raymond, Pierre; Sarrabezolles, Louise; Lucas, Didier

doi:10.1117/12.2536654

Cited by 1 publication

(1 citation statement)

References 1 publication

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This study fine-tunes the MobileNetV2 21 SSD 22 object detection model, pretrained with the COCO dataset. 19 This CNN model has already been implemented successfully during previous studies 23,24 and is particularly well adapted for embedded applications. 25 In subsection 4.5.2 additionnally a MobileNetV2 21 classifier without the SSD layer, pretrained with the ImageNet 26 dataset, has been fine-tuned for comparison reasons.…”

Section: Cnn Modelmentioning

confidence: 99%

Investigating sanitized and controlled image dataset to train deep convolutional neural networks for remote object detection on the field

Pichler

Hueber

Hennequin

2022

Electro-Optical Remote Sensing XVI

View full text Add to dashboard Cite

Performing specific object detection and recognition at the imaging sensor level, raises many technical and scientific challenges. Today state-of-the-art detection performances are obtained with Deep Convolutional Neural Network (CNN) models. However reaching the expected CNN behavior in terms of sensitivity and specificity require to master the training dataset. We explore in this paper, a new way of acquiring images of military vehicles in sanitized and controlled conditions of the laboratory in order to train a CNN to recognize the same visual signature with real vehicles in realistic outdoor situations. By combining sanitized images, counter-examples and different data augmentation techniques, our investigations aim at reducing the needs of complex outdoor image acquisition. First results demonstrate the feasibility to detect and classify, in real situations, military vehicles by exploiting only signatures from miniature models.

show abstract

Section: Cnn Modelmentioning

confidence: 99%