“Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks”の学習報告

We propose a linear-time line segment detector that gives accurate results, a controlled number of false detections, and requires no parameter tuning. This algorithm is tested and compared to state-of-the-art algorithms on a wide set of natural images.

LSD: A Fast Line Segment Detector with a False Detection Control

Aerial scene classification, which aims to automatically label an aerial image with a specific semantic category, is a fundamental problem for understanding high-resolution remote sensing imagery. In recent years, it has become an active task in the remote sensing area, and numerous algorithms have been proposed for this task, including many machine learning and data-driven approaches. However, the existing data sets for aerial scene classification, such as UC-Merced data set and WHU-RS19, contain relatively small sizes, and the results on them are already saturated. This largely limits the development of scene classification algorithms. This paper describes the Aerial Image data set (AID): a large-scale data set for aerial scene classification. The goal of AID is to advance the state of the arts in scene classification of remote sensing images. For creating AID, we collect and annotate more than 10000 aerial scene images. In addition, a comprehensive review of the existing aerial scene classification techniques as well as recent widely used deep learning methods is given. Finally, we provide a performance analysis of typical aerial scene classification and deep learning approaches on AID, which can be served as the baseline results on this benchmark.

AID: A Benchmark Data Set for Performance Evaluation of Aerial Scene Classification

Aerial scene classification, which aims to automatically label an aerial image with a specific semantic category, is a fundamental problem for understanding high-resolution remote sensing imagery. In recent years, it has become an active task in remote sensing area and numerous algorithms have been proposed for this task, including many machine learning and data-driven approaches. However, the existing datasets for aerial scene classification like UC-Merced dataset and WHU-RS19 are with relatively small sizes, and the results on them are already saturated. This largely limits the development of scene classification algorithms. This paper describes the Aerial Image Dataset (AID): a large-scale dataset for aerial scene classification. The goal of AID is to advance the state-of-the-arts in scene classification of remote sensing images. For creating AID, we collect and annotate more than ten thousands aerial scene images. In addition, a comprehensive review of the existing aerial scene classification techniques as well as recent widely-used deep learning methods is given. Finally, we provide a performance analysis of typical aerial scene classification and deep learning approaches on AID, which can be served as the baseline results on this benchmark.

AID: A Benchmark Dataset for Performance Evaluation of Aerial Scene Classification

LSD is a linear-time Line Segment Detector giving subpixel accurate results. It is designed to work on any digital image without parameter tuning. It controls its own number of false detections: on average, one false alarm is allowed per image [1]. The method is based on Burns, Hanson, and Riseman’s method [2], and uses an a contrario validation approach according to Desolneux, Moisan, and Morel’s theory [3, 4]. The version described here includes some further

/pdf/lsd-a-line-segment-detector-1bjldt7rdv.pdf

LSD: a Line Segment Detector

Images acquired by the Moderate Resolution Imaging Spectroradiometer (MODIS) aboard Terra and Aqua exhibit strong detector striping. This artifact is common to most pushbroom scanners and affects both visual interpretation and radiometric integrity of remotely sensed data. A considerable effort has been made to remove stripe noise and reduce its impact on high-level products. Despite the variety of destriping algorithms proposed in the literature, complete removal of stripes without signal distortion is yet to be overcome. In this paper, we tackle the striping issue from a variational angle. Basic statistical assumptions used in previous techniques are replaced by a much realistic geometrical consideration on the striping unidirectional variations. The resulting algorithm is tested on Aqua and Terra MODIS data contaminated with severe stripes and is shown to provide optimal qualitative and quantitative results.

Toward Optimal Destriping of MODIS Data Using a Unidirectional Variational Model

Among all methods for reconstructing missing regions in a digital image, the so-called exemplar-based algorithms are very efficient and often produce striking results. They are based on the simple idea—initially used for texture synthesis—that the unknown part of an image can be reconstructed by simply pasting samples extracted from the known part. Beyond heuristic considerations, there have been very few contributions in the literature to explain from a mathematical point of view the performances of these purely algorithmic and discrete methods. More precisely, a recent paper by Levina and Bickel [Ann. Statist., 34 (2006), pp. 1751–1773] provides a theoretical explanation of their ability to recover very well the texture, but nothing equivalent has been done so far for the recovery of geometry. Our purpose in this paper is twofold: (1) to propose well-posed variational models in the continuous domain that can be naturally associated to exemplar-based algorithms; (2) to investigate their ability to recons...

/pdf/exemplar-based-inpainting-from-a-variational-point-of-view-1wquei2buy.pdf

Exemplar-Based Inpainting from a Variational Point of View

We propose a method for the weakly supervised detection of objects in paintings. At training time, only image-level annotations are needed. This, combined with the efficiency of our multiple-instance learning method, enables one to learn new classes on-the-fly from globally annotated databases, avoiding the tedious task of manually marking objects. We show on several databases that dropping the instance-level annotations only yields mild performance losses. We also introduce a new database, IconArt, on which we perform detection experiments on classes that could not be learned on photographs, such as Jesus Child or Saint Sebastian. To the best of our knowledge, these are the first experiments dealing with the automatic (and in our case weakly supervised) detection of iconographic elements in paintings. We believe that such a method is of great benefit for helping art historians to explore large digital databases.

/pdf/weakly-supervised-object-detection-in-artworks-89dcmmq3jt.pdf

Weakly Supervised Object Detection in Artworks

Space agencies are rapidly building up massive image databases. A particularity of these databases is that they are made of images with different, but known, resolutions. In this paper, we introduce a new scheme allowing us to compare and index images with different resolutions. This scheme relies on a simplified acquisition model of satellite images and uses continuous wavelet decompositions. We establish a correspondence between scales which permits us to compare wavelet decompositions of images having different resolutions. We validate the approach through several matching and classification experiments, and we show that taking the acquisition process into account yields better results than just using scaling properties of wavelet features.

/pdf/indexing-of-satellite-images-with-different-resolutions-by-1cnn0bq947.pdf

Indexing of Satellite Images With Different Resolutions by Wavelet Features

We address the problem of computing a local orientation map in a digital image. We show that standard image gray level quantization causes a strong bias in the repartition of orientations, hindering any accurate geometric analysis of the image. In continuation, a simple dequantization algorithm is proposed, which maintains all of the image information and transforms the quantization noise in a nearby Gaussian white noise (we actually prove that only Gaussian noise can maintain isotropy of orientations). Mathematical arguments are used to show that this results in the restoration of a high quality image isotropy. In contrast with other classical methods, it turns out that this property can be obtained without smoothing the image or increasing the signal-to-noise ratio (SNR). As an application, it is shown in the experimental section that, thanks to this dequantization of orientations, such geometric algorithms as the detection of nonlocal alignments can be performed efficiently. We also point out similar improvements of orientation quality when our dequantization method is applied to aliased images.

/pdf/dequantizing-image-orientation-2e5h8zgkgq.pdf

Saïd Ladjal

Papers

Toward Optimal Destriping of MODIS Data Using a Unidirectional Variational Model

Exemplar-Based Inpainting from a Variational Point of View

Weakly Supervised Object Detection in Artworks

Indexing of Satellite Images With Different Resolutions by Wavelet Features

Dequantizing image orientation