diff --git a/ccv/poster/coffee-machine-sequence.gif b/ccv/poster/coffee-machine-sequence.gif new file mode 100644 index 0000000..3761fc9 Binary files /dev/null and b/ccv/poster/coffee-machine-sequence.gif differ diff --git a/ccv/poster/global-recall-table_vs_table-small.gif b/ccv/poster/global-recall-table_vs_table-small.gif new file mode 100644 index 0000000..6e73f3d Binary files /dev/null and b/ccv/poster/global-recall-table_vs_table-small.gif differ diff --git a/ccv/poster/method-object-discovery.png b/ccv/poster/method-object-discovery.png new file mode 100644 index 0000000..a19c1ee Binary files /dev/null and b/ccv/poster/method-object-discovery.png differ diff --git a/ccv/poster/poster-draft.svg b/ccv/poster/poster-draft.svg new file mode 100644 index 0000000..d4e15c8 --- /dev/null +++ b/ccv/poster/poster-draft.svg @@ -0,0 +1,252 @@ + + + + + + + + + + image/svg+xml + + + + + + + + + + + + + + + + + + + + Colour and depth info are combined with inhibition of return map to generate non-inhibited candidates[1]Washington dataset too simple, therefore own "Coffee machine sequence" createdInhibition of Returns with 20 candidates/frame is almost as good as 255 candidates/frame without inhibition of returnColour+Depth better than colour, depth, Alexe, Manen or Potapova + + + + + A computational framework for attentional object discovery in RGB-D videos Authors: Germán Martín García, Mircea Pavel, Simone Frintrop Presented by: Jim Martens Literature:[1]: García, Germán Martín, Mircea Pavel, and Simone Frintrop. "A computational framework for attentional object discovery in RGB-D videos." Cognitive Processing (2017): 1-14.[2] Alexe B, Deselaers T, Ferrari V (2012) Measuring the objectness of image windows. IEEE Trans Pattern Anal Mach Intell 34(11):2189–2202[3] Manén S, Guillaumin M, Van Gool L (2013) "Prime object proposals with randomized Prim’s algorithm." In: IEEE International Conference on Computer Vision (ICCV)[4] Potapova E et al. (2014) "Attention-driven object detection and segmentation ofcluttered table scenes using 2.5D symmetry." In: Proceedings of the IEEE International Conference on Robotics and Automation (ICRA) + diff --git a/ccv/poster/poster-draft_v2.svg b/ccv/poster/poster-draft_v2.svg new file mode 100644 index 0000000..66cebd7 --- /dev/null +++ b/ccv/poster/poster-draft_v2.svg @@ -0,0 +1,338 @@ + + + + + + + + + + image/svg+xml + + + + + + + + + + + + + + + + + + + + + + + + + + A computational framework for attentional object discovery in RGB-D videos Authors: Germán Martín García, Mircea Pavel, Simone Frintrop Presented by: Jim Martens Literature:[1]: García, Germán Martín, Mircea Pavel, and Simone Frintrop. "A computational framework for attentional object discovery in RGB-D videos." Cognitive Processing (2017): 1-14.[2] Alexe B, Deselaers T, Ferrari V (2012) Measuring the objectness of image windows. IEEE Trans Pattern Anal Mach Intell 34(11):2189–2202[3] Manén S, Guillaumin M, Van Gool L (2013) "Prime object proposals with randomized Prim’s algorithm." In: IEEE International Conference on Computer Vision (ICCV)[4] Potapova E et al. (2014) "Attention-driven object detection and segmentation ofcluttered table scenes using 2.5D symmetry." In: Proceedings of the IEEE International Conference on Robotics and Automation (ICRA) + Introduction object detection important for scene interpretationcomputational model proposed which combines existing work into coherent framework Methods inhibition of return mechanism that operates on spatial coordinates (using KinectFusion)object discovery: saliency for object location and segmentation to refine the boundaries projection of 2D IOR map to object candidatesranking of object candidates + Results + diff --git a/ccv/poster/precision-recall-diagrams.gif b/ccv/poster/precision-recall-diagrams.gif new file mode 100644 index 0000000..b334812 Binary files /dev/null and b/ccv/poster/precision-recall-diagrams.gif differ diff --git a/ccv/poster/successful-candidates-for-segmentation-methods.png b/ccv/poster/successful-candidates-for-segmentation-methods.png new file mode 100644 index 0000000..4e6606a Binary files /dev/null and b/ccv/poster/successful-candidates-for-segmentation-methods.png differ diff --git a/ccv/poster/system-overview.png b/ccv/poster/system-overview.png new file mode 100644 index 0000000..4226830 Binary files /dev/null and b/ccv/poster/system-overview.png differ diff --git a/ccv/poster/washington-dataset-recall_precision.gif b/ccv/poster/washington-dataset-recall_precision.gif new file mode 100644 index 0000000..98727db Binary files /dev/null and b/ccv/poster/washington-dataset-recall_precision.gif differ