DaLI: Deformation and Light Invariant Descriptor

Simo-Serra, Edgar; Torras, Carme; Moreno-Noguer, Francesc

doi:10.1007/s11263-015-0805-1

Cited by 34 publications

(31 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lastly, we evaluate our descriptors on a recent, publicly available dataset featuring challenging non-rigid deformations and very severe illumination changes [22]. The dataset consists of a series of photographs of 12 deformable ob- Figure 7: Samples from the experiments of Sec.…”

Section: Deformation and Varying Illumination Datasetmentioning

confidence: 99%

“…4.5 (top, dataset from [25]) and Sec. 4.6 (bottom, dataset from [22] jects, such as clothes and newspapers, which are subjected to four different deformation levels and four different illumination levels, i.e. 16 images per object, for a total of 192 grayscale 640×480 images.…”

Section: Deformation and Varying Illumination Datasetmentioning

confidence: 99%

“…We perform in-depth comparisons against both traditional, hand-crafted descriptors [16,27,22] as well as learned, state-of-the-art descriptors [23,29], using Precision-Recall (PR) and its area under the curve (AUC) as a metric, and demonstrate consistent gains in performance. Our descriptors also generalize very well to applications for which they were not specifically trained, demonstrating remarkable robustness against scaling, rotation, viewpoint changes, non-rigid deformations, and varying illumination.…”

Section: Introductionmentioning

confidence: 97%

“…While most descriptors, such as SIFT [16], rely on hand-crafted features [1,13,16,22,27,28,32], there has recently been interest in using machine learning algorithms to learn them from large datasets [20,23,29].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Discriminative Learning of Deep Convolutional Feature Point Descriptors

Simo-Serra

Trulls

Ferraz³

et al. 2015

2015 IEEE International Conference on Computer Vision (ICCV)

Self Cite

737

650

View full text Add to dashboard Cite

Deep learning has revolutionalized image-level tasks such as classification, but patch-level tasks, such as correspondence, still rely on hand-crafted features, e.g. SIFT. In this paper we use Convolutional Neural Networks (CNNs) to learn discriminant patch representations and in particular train a Siamese network with pairs of (non-)corresponding patches. We deal with the large number of potential pairs with the combination of a stochastic sampling of the training set and an aggressive mining strategy biased towards patches that are hard to classify.By using the L 2 distance during both training and testing we develop 128-D descriptors whose euclidean distances reflect patch similarity, and which can be used as a drop-in replacement for any task involving SIFT. We demonstrate consistent performance gains over the state of the art, and generalize well against scaling and rotation, perspective transformation, non-rigid deformation, and illumination changes. Our descriptors are efficient to compute and amenable to modern GPUs, and are publicly available.

show abstract

Section: Deformation and Varying Illumination Datasetmentioning

confidence: 99%

Section: Deformation and Varying Illumination Datasetmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 97%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Discriminative Learning of Deep Convolutional Feature Point Descriptors

Simo-Serra

Trulls

Ferraz³

et al. 2015

2015 IEEE International Conference on Computer Vision (ICCV)

Self Cite

737

650

View full text Add to dashboard Cite

show abstract

“…We show quantitative results that demonstrate a large improvement over the widely used Gaussian diffusion models. Furthermore, it is straightforward to extend existing 3D human pose estimation algorithms [11,12] to tracking using the proposed prior using stronger image features [14,15].…”

Section: Discussionmentioning

confidence: 99%

Lie Algebra-Based Kinematic Prior for 3D Human Pose Tracking

Simo-Serra

Torras

Moreno-Noguer

2015

2015 14th IAPR International Conference on Machine Vision Applications (MVA)

Self Cite

View full text Add to dashboard Cite

We propose a novel kinematic prior for 3D human pose tracking that allows predicting the position in subsequent frames given the current position. We first define a Riemannian manifold that models the pose and extend it to also be able to represent the kinematics. We then learn a joint Gaussian mixture model of both the human pose and the kinematics on this manifold. Finally by conditioning the kinematics on the pose we are able to obtain a distribution of poses for subsequent frames that which can be used as a reliable prior in 3D human pose tracking. Our model scales well to large amounts of data and can be sampled at over 100,000 samples/second. We show it outperforms the widely used Gaussian diffusion model on the challenging Human3.6M dataset.

show abstract

Deep Self-correlation Descriptor for Dense Cross-Modal Correspondence

Kim

Min

Lin

et al. 2016

Computer Vision – ECCV 2016

View full text Add to dashboard Cite

We present a novel descriptor, called deep self-convolutional activations (DeSCA), designed for establishing dense correspondences between images taken under different imaging modalities, such as different spectral ranges or lighting conditions. Motivated by descriptors based on local self-similarity (LSS), we formulate a novel descriptor by leveraging LSS in a deep architecture, leading to better discriminative power and greater robustness to non-rigid image deformations than stateof-the-art cross-modality descriptors. The DeSCA first computes selfconvolutions over a local support window for randomly sampled patches, and then builds self-convolution activations by performing an average pooling through a hierarchical formulation within a deep convolutional architecture. Finally, the feature responses on the self-convolution activations are encoded through a spatial pyramid pooling in a circular configuration. In contrast to existing convolutional neural networks (CNNs) based descriptors, the DeSCA is training-free (i.e., randomly sampled patches are utilized as the convolution kernels), is robust to cross-modal imaging, and can be densely computed in an efficient manner that significantly reduces computational redundancy. The state-of-the-art performance of DeSCA on challenging cases of cross-modal image pairs is demonstrated through extensive experiments.

show abstract

DaLI: Deformation and Light Invariant Descriptor

Cited by 34 publications

References 49 publications

Discriminative Learning of Deep Convolutional Feature Point Descriptors

Discriminative Learning of Deep Convolutional Feature Point Descriptors

Lie Algebra-Based Kinematic Prior for 3D Human Pose Tracking

Deep Self-correlation Descriptor for Dense Cross-Modal Correspondence

Contact Info

Product

Resources

About