Code and datasets

Sediment tracking

In collaboration with ETNA group of Irstea Grenoble, we recorded video sequences to study sediment transport. Experiments consist in a two-dimensional steep channel where mixtures of two-size spherical glass beads are entrained by a turbulent supercritical water flow with a mobile bed.

The aim is to track all beads over time to obtain trajectories, particle velocities and concentrations 1, for studying bedload granular rheology, size segregation and associated morphology.

The code implementing the tracking algorithms 2 is available here. Two datasets are provided:

Experimental sequence (view, download, full dataset)
This is a 1000-frame sequence recorded at 130 fps with approximately 400 beads per frame (about 300 coarse and 100 small beads). The image resolution is 1280x320.
Numerical sequence (view, download, full dataset)
This 10,000 frame sequence was generated thanks to a model developed at Irstea based on a coupled fluid discrete element method.

AVA symmetry

From DPChallenge photo contest website, Murray et al. 3 introduces different annotations (aesthetic, semantic and photographic style) for more than 250,000 images for Aesthetic Visual Analysis “AVA”. From the following photography challenges, we labeled global-axis symmetry groundtruth 4 for 253 out of 878 images: (1) five challenges of “Reflections Without Mirrors”: images containing bilateral representation without using mirror, (2) three challenges of “Symmetry”: photographs composing symmetrical balance. These images are selected to neglect unclear instances of ambiguity symmetry, and to represent many comparison cases (non-centering viewpoint, perspective view, blurring reflection, etc.) for detection algorithms.

The dataset is publicly available through GitHub.

  1. Online multi-model particle filter-based tracking to study bedload transport, H. Lafaye de Micheaux, C. Ducottet, P. Frey, ICIP 2016, IEEE International Conference on Image Processing, pp. 3489-3493, 2016.↩

  2. Multi-model particle filter-based tracking with switching dynamical state to study bedload transport, H. Lafaye de Micheaux, C. Ducottet, P. Frey, Machine Vision and Applications, Vol. 29(5), pp. 735-747, 2018.↩

  3. Murray, N., Marchesotti, L., Perronnin, F.: AVA: A large-scale database for aesthetic visual analysis. In: Computer Vision and Pattern Recognition (CVPR), 2012 IEEE Conference on. pp. 2408–2415, 2012.↩

  4. Global Bilateral Symmetry Detection Using Multiscale Mirror Histograms, M. Elawady, C. Barat, C. Ducottet, P. Colantoni, In International Conference on Advanced Concepts for Intelligent Vision Systems, pp. 14-24, 2016.↩