Test-Time Adaptation for Optical Flow Estimation Using Motion Vectors.

Due to the prohibitive cost as well as technical challenges in annotating ground-truth optical flow for large-scale realistic video datasets, the existing deep learning models for optical flow estimation mostly rely on synthetic data for training, which in turn may lead to significant performance degradation under test-data distribution shift in real-world environments. In this work, we propose the methodology to tackle this important problem. We design a self-supervised learning task for adjusting the optical flow estimation model at test time. We exploit the fact that most videos are stored in compressed formats, from which compact information on motion, in the form of motion vectors and residuals, can be made readily available. We formulate the self-supervised task as motion vector prediction, and link this task to optical flow estimation. To the best of our knowledge, our Test-Time Adaption guided with Motion Vectors (TTA-MV), is the first work to perform such adaptation for optical flow. The experimental results demonstrate that TTA-MV can improve the generalization capability of various well-known deep learning methods for optical flow estimation, such as FlowNet, PWCNet, and RAFT.

A new procedure, free from human assessment that automatically grades some facial skin structural signs. Comparison with assessments by experts, using referential atlases of skin ageing.

OBJECTIVE: To develop an automatic system that grades the severity of facial signs through 'selfies' pictures taken by women of different ages and ethnics. METHODS: 1140 women from three ethnics (African-American, Asian, Caucasian), of different ages (18-80 years old), took 'selfies' by high resolution smartphones cameras under different conditions of lighting or facial expressions. A dedicated software, was developed, based on a Convolutional Neural Network (CNN) that integrates training data from referential Skin Aging Atlases. The latter allows to an immediate quantification of the severity of nine facial signs according to the ethnicity declared by the subject. These automatic grading were confronted to those assessed by 12 trained experts and dermatologists either on 'selfies' pictures or in live conditions on a smaller cohort of women. RESULTS: The system appears weakly influenced by lighting conditions or facial expressions (coefficients of variations ranging 10-13% for most signs) and leads to global agreements with experts' assessments, even showing a better reproducibility on some facial signs. CONCLUSION: This automatic scoring system, still in development, seems offering a new quantitative approach in the quantified description of facial signs, independent from human vision, in many applications, being individual, cosmetic oriented or dermatological with regard to the follow-up of medical anti-ageing corrective strategies.

OBJECTIF: De développer un système automatique qui quantifie la sévérité de certains signes du visage à partir de photographies de type 'selfies' pris par des femmes d'origine ethnique et d'âge différents. MÉTHODES: 1140 femmes de trois ethnies différentes (Afro-Américaines, Asiatiques, Caucasiennes), d'âges différents (18-80 ans) ont pris des selfies sous différentes conditions d'éclairage et d'expressions faciales. Un logiciel dédié a été développé, basé sur un réseau de convolution neuronal et intégrant les données d'annotations utilisant les Atlas de Vieillissement Cutané. Ce système quantifie immédiatement la sévérité de 9 signes faciaux selon l'ethnie déclarée par le sujet. Ces scores ont été confrontés à ceux de 12 experts et dermatologistes soit à partir des 'selfies' ou en conditions réelles sur un groupe plus restreint de femmes. RÉSULTATS: Le système apparaît faiblement influencé par les conditions d'éclairage et les expressions faciales (coefficients de variation de l'ordre de 10-13%) et conduit à des valeurs comparables de celles des experts, voire même de meilleure reproductibilité dans certains cas. CONCLUSION: Ce système de scorage automatique, encore en développement, semble offrir une nouvelle approche dans la description quantitative de signes du visage, indépendante de l'œil humain, dans de nombreuses applications, comme la personnalisation, à visée cosmétique ou dermatologique, dans le suivi de certaines stratégies médicales de l'antivieillissement cutané.

Mean q-ball strings obtained by constrained procrustes analysis with point sliding.

The idea underpinning the work we present herein is to design robust and objective tools for brain white matter (WM) morphometry. We focus on WM tracts, and propose to represent them by their mean lines, to which we associate the attributes derived from high-angular resolution diffusion imaging (HARDI). The definition of the tract mean line derives directly from the geometry of the tract fibres. We determine the fibre point correspondences and impact factors of individual fibres, upon which we estimate average HARDI models along the tract mean lines. This way we obtain a compact tract representation that exploits all the available information, and is at the same time free of the outlier influence and undesired tract edge effects.

New diffusion phantoms dedicated to the study and validation of high-angular-resolution diffusion imaging (HARDI) models.

We present new diffusion phantoms dedicated to the study and validation of high-angular-resolution diffusion imaging (HARDI) models. The phantom design permits the application of imaging parameters that are typically employed in studies of the human brain. The phantoms were made of small-diameter acrylic fibers, chosen for their high hydrophobicity and flexibility that ensured good control of the phantom geometry. The polyurethane medium was filled under vacuum with an aqueous solution that was previously degassed, doped with gadolinium-tetraazacyclododecanetetraacetic acid (Gd-DOTA), and treated by ultrasonic waves. Two versions of such phantoms were manufactured and tested. The phantom's applicability was demonstrated on an analytical Q-ball model. Numerical simulations were performed to assess the accuracy of the phantom. The phantom data will be made accessible to the community with the objective of analyzing various HARDI models.

Voxel-based modeling and quantification of the proximal femur using inter-subject registration of quantitative CT images.

We have developed a general framework which employs quantitative computed tomography (QCT) imaging and inter-subject image registration to model the three-dimensional structure of the hip, with the goal of quantifying changes in the spatial distribution of bone as it is affected by aging, drug treatment or mechanical unloading. We have adapted rigid and non-rigid inter-subject registration techniques to transform groups of hip QCT scans into a common reference space and to construct composite proximal femoral models. We have applied this technique to a longitudinal study of 16 astronauts who on average, incurred high losses of hip bone density during spaceflights of 4-6 months on the International Space Station (ISS). We compared the pre-flight and post-flight composite hip models, and observed the gradients of the bone loss distribution. We performed paired t-tests, on a voxel by voxel basis, corrected for multiple comparisons using false discovery rate (FDR), and observed regions inside the proximal femur that showed the most significant bone loss. To validate our registration algorithm, we selected the 16 pre-flight scans and manually marked 4 landmarks for each scan. After registration, the average distance between the mapped landmarks and the corresponding landmarks in the target scan was 2.56 mm. The average error due to manual landmark identification was 1.70 mm.