We propose to tackle the problem of RGB-D image disocclusion inpainting when synthesizing new views of a scene by changing its viewpoint. Indeed, such a process creates holes both in depth and color images. First, we propose a novel algorithm to perform depth-map disocclusion inpainting. Our intuitive approach works particularly well for recovering the lost structures of the objects and to inpaint the depth-map in a geometrically plausible manner. Then, we propose a depth-guided patch-based inpainting method to fill-in the color image. Depth information coming from the reconstructed depth-map is added to each key step of the classical patch-based algorithm from Criminisi et al. in an intuitive manner. Relevant comparisons to the state-of-the-art inpainting methods for the disocclusion inpainting of both depth and color images are provided and illustrate the effectiveness of our proposed algorithms.

In this paper we propose an approach to inpaint holes in depth maps that appear when synthesizing virtual views from a RGB-D scenes. Based on a superpixel oversegmentation of both the original and synthesized views, the proposed approach efficiently deals with many occlusion situations where most of previous approaches fail. The use of superpixels makes the algorithm more robust to inaccurate depth maps, while giving an efficient way to model the image. Extensive comparisons to relevant state-of-the-art methods show that our approach outperforms qualitatively and quantitavely these existing approaches.

Diffusion methods have proven their efficiency for tasks such as semi-supervised segmentation. The introduction of patches as a part of their speed function allows to deal with textured images. However, the computational burden with such variants stays too important for low-level tasks. In this paper, we propose a multivalued color-based potential function that partly alleviates this flaw. It allows to efficiently perform semi-supervised segmentation of natural and textured images.

This paper proposes a technical review of exemplar-based inpainting approaches with a particular focus on greedy methods. Several comparative and illustrative experiments are provided to deeply explore and enlighten these methods, and to have a better understanding on the state-of-the-art improvements of these approaches. From this analysis, three improvements over Criminisi et al. algorithm are then presented and detailed: 1) a tensor-based data term for a better selection of pixel candidates to fill in; 2) a fast patch lookup strategy to ensure a better global coherence of the reconstruction; and 3) a novel fast anisotropic spatial blending algorithm that reduces typical block artifacts using tensor models. Relevant comparisons with the state-of-the-art inpainting methods are provided that exhibit the effectiveness of our contributions.

In this paper we propose to use the Eikonal equation on graphs for generalized data clustering. We introduce a new potential function that favors the creation of homogeneous clusters together with an iterative algorithm that place seeds vertices at smart locations. Oversegmentation application shows the effectiveness of our approach and gives results comparable to the state-of-the-art methods.

In this paper, we propose two major improvements to the exemplar-based image inpainting algorithm, initially formulated by Criminisi et al. [1]. First, we introduce a structure-tensor-based data-term for a better selection of pixel candidates to fill in based on priority. Then, we propose a new lookup heuristic in order to locate the best source patches to copy/paste to these targeted points. These two contributions clearly make the inpainting algorithm reconstruct more geometrically coherent images, as well as speed up the process drastically. We illustrate the great performances of our approach compared to existing state-of-the-art methods.

We present a low resolution face recognition technique based on a Convolutional Neural Network approach. The network is trained to reconstruct a reference per subject image. In classical feature-based approaches, a first stage of features extraction is followed by a classification to perform the recognition. In classical Convolutional Neural Network approaches, features extraction stages are stacked (interlaced with pooling layers) with classical neural layers on top to form the complete architecture of the network. This paper addresses two questions: 1. Does a pretraining of the filters in an unsupervised manner improve the recognition rate compared to the one with filters learned in a purely supervised scheme? 2. Is there an advantage of pretraining more than one feature extraction stage? We show particularly that a refinement of the filters during the supervised training improves the results.