Even though data annotation is extremely important for interpretability, research, and development of artificial intelligence solutions, annotating data remains costly. Research efforts such as active learning or few-shot learning alleviate the cost by increasing sample efficiency, yet the problem of annotating data more quickly has received comparatively little attention. Leveraging a predictor has been shown to reduce annotation cost in practice but has not been theoretically considered. We ask the following question: to annotate a binary classification dataset with N samples, can the annotator answer less than N yes/no questions? Framing this question-and-answer (Q&A) game as an optimal encoding problem, we find a positive answer given by the Huffman encoding of the possible labelings. Unfortunately, the algorithm is computationally intractable even for small dataset sizes. As a practical method, we propose to minimize a cost function a few steps ahead, similarly to lookahead minimization in optimal control. This solution is analyzed, compared with the optimal one, and evaluated using several synthetic and real-world datasets. The method allows a significant improvement (23-86%) in the annotation efficiency of real-world datasets.

Self-supervised representation learning based on Contrastive Learning (CL) has been the subject of much attention in recent years. This is due to the excellent results obtained on a variety of subsequent tasks (in particular classification), without requiring a large amount of labeled samples. However, most reference CL algorithms (such as SimCLR and MoCo, but also BYOL and Barlow Twins) are not adapted to pixel-level downstream tasks. One existing solution known as PixPro proposes a pixel-level approach that is based on filtering of pairs of positive/negative image crops of the same image using the distance between the crops in the whole image. We argue that this idea can be further enhanced by incorporating semantic information provided by exogenous data as an additional selection filter, which can be used (at training time) to improve the selection of the pixel-level positive/negative samples. In this paper we will focus on the depth information, which can be obtained by using a depth estimation network or measured from available data (stereovision, parallax motion, LiDAR, etc.). Scene depth can provide meaningful cues to distinguish pixels belonging to different objects based on their depth. We show that using this exogenous information in the contrastive loss leads to improved results and that the learned representations better follow the shapes of objects. In addition, we introduce a multi-scale loss that alleviates the issue of finding the training parameters adapted to different object sizes. We demonstrate the effectiveness of our ideas on the Breakout Segmentation on Borehole Images where we achieve an improvement of 1.9% over PixPro and nearly 5% over the supervised baseline. We further validate our technique on the indoor scene segmentation tasks with ScanNet and outdoor scenes with CityScapes (1.6% and 1.1% improvement over PixPro respectively).

We tackle the problem of texture inpainting where the input images are textures with missing values along with masks that indicate the zones that should be generated. Many works have been done in image inpainting with the aim to achieve global and local consistency. But these works still suffer from limitations when dealing with textures. In fact, the local information in the image to be completed needs to be used in order to achieve local continuities and visually realistic texture inpainting. For this, we propose a new segmentor discriminator that performs a patch-wise real/fake classification and is supervised by input masks. During training, it aims to locate the generated parts (which we will call fake parts in consistency with the GAN framework), thus making the difference between real and fake patches given one image. We tested our approach on the publicly available DTD dataset, as well as Electo-Magnetic borehole images dataset and showed that it achieves state-of-the-art performances and better deals with local consistency than existing methods.

Borehole images are often misaligned due to a depth offset in the acquisition sensors of a wireline tool. We propose an algorithm that identifies matching feature points and aligns the image. One of the main challenges is that imaging pads have no azimuthal overlap. This has been solved by extrapolating the pixels on the boundary to create a synthetic overlap that facilitates feature matching. Feature matching is implemented in two stages: first by finding a local correspondence among feature points and second by performing a global minimization with additional regularization constraints on the estimated shifts. The method has been successfully used to align several imaging logs in less than 1 minute whereas a manual alignment would take 6 hours.

Natural fractures may greatly affect wellbore stability and fluid flow in oil reservoirs. The detection of fracture heterogeneities-also named fracture traces-on electrical image logs is not yet efficiently automated and consequently requires long and tedious work from geologists. In this article, we propose an efficient fracture trace extraction algorithm that produces fast, efficient and repeatable results. Morphology operations are first applied to detect high contrast traces while controlling their geometry. A priori information about the geometry of sedimentary surfaces is then used to compute the main fracture orientation. Finally, a clustering algorithm is performed for grouping the extracted traces and identifying those that correspond to natural fractures.